[Feature] Support PointRCNN RPN and RCNN module #1022

wHao-Wu · 2021-11-01T03:55:58Z

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Support PointRCNN RPN and RCNN module to support the complete PointRCNN detector.

Modification

Support PointRCNN RPN and RCNN module.

BC-breaking (Optional)

Does the modification introduce changes that break the back-compatibility of the downstream repos?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Checklist

Pre-commit or other linting tools are used to fix the potential lint issues.
The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
If the modification has potential influence on downstream projects, this PR should be tested with downstream projects.
The documentation has been modified accordingly, like docstring or example tutorials.

mmdet3d/models/dense_heads/point_rpn_head.py

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py

mmdet3d/models/dense_heads/point_rpn_head.py

configs/_base_/models/point_rcnn.py

configs/pointrcnn/README.md

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py

mmdet3d/models/dense_heads/ssd_3d_head.py

mmdet3d/models/dense_heads/point_rpn_head.py

CLAassistant · 2021-12-01T07:49:04Z

All committers have signed the CLA.

mmdet3d/models/dense_heads/point_rpn_head.py

mmdet3d/models/detectors/pointrcnn.py

Machine97 · 2021-12-02T12:10:03Z

Thank you for your contribution. When I try to train PointRCNN as you did, it get stuck in the first epoch during training, and no error warning. Can you help me to solve this problem?

wHao-Wu · 2021-12-03T02:37:04Z

Hi, @Machine97
Thanks for your attention. For training PointRCNN successfully, you may meet the prerequisites as follows:

Please pull and install the latest mmcv and mmdetection.
If you generate KITTI data using the master branch, please regenerate KITTI data using the v1.0.0.dev0 branch.
Please install the mmdetection3d instructed by here with the v1.0.0.dev0 branch instead of the master branch.

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py

ZwwWayne · 2021-12-07T09:03:21Z

Thank you for your contribution. When I try to train PointRCNN as you did, it get stuck in the first epoch during training, and no error warning. Can you help me to solve this problem?

There is indeed an issue in the loss calculation when there is no positive sample. We will fix that in this PR.

configs/_base_/models/point_rcnn.py

mmdet3d/models/dense_heads/point_rpn_head.py

mmdet3d/models/detectors/point_rcnn.py

mmdet3d/models/dense_heads/point_rpn_head.py

Machine97 · 2021-12-14T07:51:28Z

Hi, all. Thank you for your contribution. I want to ask have you complished the PointRCNN and what's the detailed results?

Machine97 · 2021-12-14T08:29:38Z

Hi, @Machine97 Thanks for your attention. For training PointRCNN successfully, you may meet the prerequisites as follows:

Please pull and install the latest mmcv and mmdetection.

If you generate KITTI data using the master branch, please regenerate KITTI data using the v1.0.0.dev0 branch.

Please install the mmdetection3d instructed by here with the v1.0.0.dev0 branch instead of the master branch.

Thank you for your reply. I'm sorry to reply so late. I want to ask if I build an image according to the v1.0.0.dev0 branch Dockerfile, then train the PointRCNN. Is that ok?

Machine97 · 2021-12-15T02:09:16Z

Hi, @Machine97 Thanks for your attention. For training PointRCNN successfully, you may meet the prerequisites as follows:

Please pull and install the latest mmcv and mmdetection.

If you generate KITTI data using the master branch, please regenerate KITTI data using the v1.0.0.dev0 branch.

Please install the mmdetection3d instructed by here with the v1.0.0.dev0 branch instead of the master branch.

Thank you for your reply. I'm sorry to reply so late. I want to ask if I build an image according to the v1.0.0.dev0 branch Dockerfile, then train the PointRCNN. Is that ok?

I tried to train PointRCNN like this, but it get stuck at the begining during training and no error warning. I'm confused about this. The following picture is the stucked position.

Machine97 · 2021-12-15T02:31:59Z

Now this error has occurred.

Machine97 · 2021-12-15T03:36:56Z

@wHao-Wu @ZwwWayne @Tai-Wang
Another question. I change an enviroment and train the PointRCNN sucessfully. But in the processing of training, it occur the following error. It seems that variable grad_norm is unstable. I don't know what the purpose of the variable grad_norm is. Can you give me some advice to solve this problem? Thank you so much.

Machine97 · 2021-12-15T07:49:29Z

@wHao-Wu @ZwwWayne @Tai-Wang Another question. I change an enviroment and train the PointRCNN sucessfully. But in the processing of training, it occur the following error. It seems that variable grad_norm is unstable. I don't know what the purpose of the variable grad_norm is. Can you give me some advice to solve this problem? Thank you so much.

@wHao-Wu @CLAassistant @ZwwWayne @Tai-Wang Hi, all.
In order to find the reason for this error, I add some code as shown in the figure below in /mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py.

Then, after training, the results shown in the figure below appear. The error seems to come from the variable 'cls_score', can you give me some advice to solve this problem?

wHao-Wu · 2021-12-16T07:58:50Z

Hi, @Machine97 Thanks for your attention. For training PointRCNN successfully, you may meet the prerequisites as follows:

Please pull and install the latest mmcv and mmdetection.

If you generate KITTI data using the master branch, please regenerate KITTI data using the v1.0.0.dev0 branch.

Please install the mmdetection3d instructed by here with the v1.0.0.dev0 branch instead of the master branch.

Thank you for your reply. I'm sorry to reply so late. I want to ask if I build an image according to the v1.0.0.dev0 branch Dockerfile, then train the PointRCNN. Is that ok?

Hi, @Machine97
Thanks for your attention and sorry for the late reply. You can build an image of v1.0.0.dev0 branch based on the Dockerfile here.

wHao-Wu · 2021-12-16T08:00:49Z

@wHao-Wu @ZwwWayne @Tai-Wang Another question. I change an enviroment and train the PointRCNN sucessfully. But in the processing of training, it occur the following error. It seems that variable grad_norm is unstable. I don't know what the purpose of the variable grad_norm is. Can you give me some advice to solve this problem? Thank you so much.

Could you please post your training log here or issue?

Machine97 · 2021-12-16T08:14:04Z

@wHao-Wu @ZwwWayne @Tai-Wang Another question. I change an enviroment and train the PointRCNN sucessfully. But in the processing of training, it occur the following error. It seems that variable grad_norm is unstable. I don't know what the purpose of the variable grad_norm is. Can you give me some advice to solve this problem? Thank you so much.

Could you please post your training log here or issue?

Hi, thank you so much for your reply.
I changed the learning rate from 0.001 to 0.0005 and it can train normally so far. By the way, I train the PointRCNN followed the configs in configs/point_rcnn/point_rcnn_2x8_kitti-3d-3classes.py with train+val data and modified the batchsize from 2 to 3, modified the lr from 0.001 to 0.0005, the rest keep as origin.
It has trained to the 12th epoch so far, the eval results are shown in the figure below. The accuracy is relatively low, is this normal? I use train+val data to train and val data to test.

wHao-Wu · 2021-12-17T05:19:46Z

@Machine97
It must be something wrong in this setting. In my case, I can get 66.6342 for Car, 41.8122 for Pedestrian, and 61.7294 for Cyclist. We have supported PointRCNN benchmark on the KITTI dataset. You can refer to here to find the log file for more details. Next, we will try to reproduce your problem. Thanks for your feedback.

Machine97 · 2021-12-17T11:28:18Z

@Machine97 It must be something wrong in this setting. In my case, I can get 66.6342 for Car, 41.8122 for Pedestrian, and 61.7294 for Cyclist. We have supported PointRCNN benchmark on the KITTI dataset. You can refer to here to find the log file for more details. Next, we will try to reproduce your problem. Thanks for your feedback.

@wHao-Wu
Thank you for your reply. This is my train log.
20211217_104132.log
I trained on 4GPUs. In the train log, the AP is very low and we can find that 'loss_cls' are gradually increasing. It must be something wrong. I'm so confused about it.

Machine97 · 2021-12-18T09:26:05Z

@wHao-Wu
Hi, could you please help me to find out where the mistake happened?

wHao-Wu · 2021-12-20T02:57:51Z

@Machine97 It must be something wrong in this setting. In my case, I can get 66.6342 for Car, 41.8122 for Pedestrian, and 61.7294 for Cyclist. We have supported PointRCNN benchmark on the KITTI dataset. You can refer to here to find the log file for more details. Next, we will try to reproduce your problem. Thanks for your feedback.

@wHao-Wu Thank you for your reply. This is my train log. 20211217_104132.log I trained on 4GPUs. In the train log, the AP is very low and we can find that 'loss_cls' are gradually increasing. It must be something wrong. I'm so confused about it.

@Machine97
Sorry for the late reply because of the time cost for issue-reproducing. I found that you haven't changed the LR to 0.0005 from your training log (the LR still keeps to 0.001). Then, I used your setting to reproduce your problem and cannot find any problem when training PointRCNN model. Detailed can be found here. Besides, I also rerun PointRCNN with LR=0.0005 and GPUs=4 and the model also got normal results.

There are some possible problems:

Please update your KITTI data with v1.0.0.dev0 if you use the master branch to generate KITTI data before.
Please pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again.
If possible, please pull the latest mmcv, mmdetection.

Machine97 · 2021-12-20T03:01:27Z

@Machine97 It must be something wrong in this setting. In my case, I can get 66.6342 for Car, 41.8122 for Pedestrian, and 61.7294 for Cyclist. We have supported PointRCNN benchmark on the KITTI dataset. You can refer to here to find the log file for more details. Next, we will try to reproduce your problem. Thanks for your feedback.

@wHao-Wu Thank you for your reply. This is my train log. 20211217_104132.log I trained on 4GPUs. In the train log, the AP is very low and we can find that 'loss_cls' are gradually increasing. It must be something wrong. I'm so confused about it.

@Machine97 Sorry for the late reply because of the time cost for issue-reproducing. I found that you haven't changed the LR to 0.0005 from your training log (the LR still keeps to 0.001). Then, I used your setting to reproduce your problem and cannot find any problem when training PointRCNN model. Detailed can be found here. Besides, I also rerun PointRCNN with LR=0.0005 and GPUs=4 and the model also got normal results.

There are some possible problems:

Please update your KITTI data with v1.0.0.dev0 if you use the master branch to generate KITTI data before.

Please pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again.

If possible, please pull the latest mmcv, mmdetection.

Thank you so much for your reply. I will try according to your advice and give you a feedback as soon as possible.

Machine97 · 2021-12-20T10:44:35Z

@Machine97 It must be something wrong in this setting. In my case, I can get 66.6342 for Car, 41.8122 for Pedestrian, and 61.7294 for Cyclist. We have supported PointRCNN benchmark on the KITTI dataset. You can refer to here to find the log file for more details. Next, we will try to reproduce your problem. Thanks for your feedback.

@wHao-Wu Thank you for your reply. This is my train log. 20211217_104132.log I trained on 4GPUs. In the train log, the AP is very low and we can find that 'loss_cls' are gradually increasing. It must be something wrong. I'm so confused about it.

@Machine97 Sorry for the late reply because of the time cost for issue-reproducing. I found that you haven't changed the LR to 0.0005 from your training log (the LR still keeps to 0.001). Then, I used your setting to reproduce your problem and cannot find any problem when training PointRCNN model. Detailed can be found here. Besides, I also rerun PointRCNN with LR=0.0005 and GPUs=4 and the model also got normal results.

There are some possible problems:

Please update your KITTI data with v1.0.0.dev0 if you use the master branch to generate KITTI data before.

Please pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again.

If possible, please pull the latest mmcv, mmdetection.

@wHao-Wu
Hi, according to your suggestion, I took the following steps:
1. Build an image of v1.0.0.dev0 branch based on the Dockerfile https://github.com/open-mmlab/mmdetection3d/blob/v1.0.0.dev0/docker/Dockerfile.
2. Update my KITTI data with v1.0.0.dev0 by running /tools/create_data.py file.
3. Pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again.
However, there still be the same mistakes as before. I think this should not be an environment problem. I want to ask what's the difference about the way of KITTI data generation in master and v1.0.0.dev0.

Machine97 · 2021-12-20T11:18:14Z

@wHao-Wu
Hi, I carefully compared the log file you provided with the latest code from v1.0.0.dev0 branch, there are some differences in /configs/base/models/point_rcnn.py, like 'enlarge_width=0.1' which doesn't appear in the log file you provided. So, would you please pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again?
I look forward to your reply as soon as possible, and appreciate it.

Machine97 · 2021-12-20T12:04:21Z

@wHao-Wu
Hi, if I change the parameters in /configs/base/models/point_rcnn.py file as the log file you provided. It will appear another error showed in the following picture.

Machine97 · 2021-12-23T05:39:58Z

@wHao-Wu
Hi, I'm so sorry to bother you again. Have you tried to pull the latest code from v1.0.0.dev0 branch and rerun the PointRCNN again?

* [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo * fix typo * fix typo * support dgcnn * Change cam rot_3d_in_axis (#906) * [Doc] Add coord sys tutorial pic and change links to dev branch (#912) * Modify link branch and add pic * Fix pic * [Feature] add kitti AP40 evaluation metric (v1.0.0.dev0) (#927) * Add citation (#901) * [Feature] Add python3.9 in CI (#900) * Add python3.0 in CI * Add python3.0 in CI * Bump to v0.17.0 (#898) * Update README.md * Update README_zh-CN.md * Update version.py * Update getting_started.md * Update getting_started.md * Update changelog.md * Remove "recent" in the news * Remove "recent" in the news * Fix comments * [Docs] Fix the version of sphinx (#902) * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * add AP40 * add unitest * add unitest * seperate AP11 and AP40 * fix some typos Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Feature] add smoke backbone neck (#939) * add smoke detecotor and it's backbone and neck * typo fix * fix typo * add docstring * fix typo * fix comments * fix comments * fix comments * fix typo * fix typo * fix * fix typo * fix docstring * refine feature * fix typo * use Basemodule in Neck * [Refactor] Refactor the transformation from image to camera coordinates (#938) * Refactor points_img2cam * Refine docstring * Support array converter and add unit tests * [Feature] FCOS3D BBox Coder (#940) * FCOS3D BBox Coder * Add unit tests * Change the value from long to float/double * Rename bbox_out as bbox * Add comments to forward returns * [Feature] PGD BBox Coder (#948) * Support PGD BBox Coder * Refine docstring * [Feature] Support Uncertain L1 Loss (#950) * Add uncertain l1 loss and its unit tests * Remove mmcv.jit and refine docstrings * [Fix] Fix visualization in KITTI dataset (#956) * fix bug to support kitti vis * fix * [Feature] Add SMOKE Head and Coder(v1.0.0.dev0) (#959) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * [Enhance] Change Groupfree3D config (#855) * All mods * PointSample * PointSample * [Doc] Add tutorials/data_pipeline Chinese version (#827) * [Doc] Add tutorials/data_pipeline Chinese version * refine doc * Use the absolute link * Use the absolute link Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add Chinese doc for `scannet_det.md` (#836) * Part * Complete * Fix comments * Fix comments * [Doc] Add Chinese doc for `waymo_det.md` (#859) * Add complete translation * Refinements * Fix comments * Fix a minor typo Co-authored-by: Tai-Wang <tab_wang@outlook.com> * Remove 2D annotations on Lyft (#867) * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [fix] fix typos (#872) * Fix 3 unworking configs (#882) * [Fix] Fix `index.rst` for Chinese docs (#873) * Fix index.rst for zh docs * Change switch language * [Fix] Centerpoint head nested list transpose (#879) * FIX Transpose nested lists without Numpy * Removed unused Numpy import * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Doc] Add nuscenes_det.md Chinese version (#854) * add nus chinese doc * add nuScenes Chinese doc * fix typo * fix typo * fix typo * fix typo * fix typo * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * [Enhance] Add benchmark regression script (#808) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo * fix typo * fix typo * support dgcnn * Change cam rot_3d_in_axis (#906) * [Doc] Add coord sys tutorial pic and change links to dev branch (#912) * Modify link branch and add pic * Fix pic * [Feature] add kitti AP40 evaluation metric (v1.0.0.dev0) (#927) * Add citation (#901) * [Feature] Add python3.9 in CI (#900) * Add python3.0 in CI * Add python3.0 in CI * Bump to v0.17.0 (#898) * Update README.md * Update README_zh-CN.md * Update version.py * Update getting_started.md * Update getting_started.md * Update changelog.md * Remove "recent" in the news * Remove "recent" in the news * Fix comments * [Docs] Fix the version of sphinx (#902) * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * add AP40 * add unitest * add unitest * seperate AP11 and AP40 * fix some typos Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Feature] add smoke backbone neck (#939) * add smoke detecotor and it's backbone and neck * typo fix * fix typo * add docstring * fix typo * fix comments * fix comments * fix comments * fix typo * fix typo * fix * fix typo * fix docstring * refine feature * fix typo * use Basemodule in Neck * [Refactor] Refactor the transformation from image to camera coordinates (#938) * Refactor points_img2cam * Refine docstring * Support array converter and add unit tests * [Feature] FCOS3D BBox Coder (#940) * FCOS3D BBox Coder * Add unit tests * Change the value from long to float/double * Rename bbox_out as bbox * Add comments to forward returns * [Feature] PGD BBox Coder (#948) * Support PGD BBox Coder * Refine docstring * [Feature] Support Uncertain L1 Loss (#950) * Add uncertain l1 loss and its unit tests * Remove mmcv.jit and refine docstrings * [Fix] Fix visualization in KITTI dataset (#956) * fix bug to support kitti vis * fix * add smoke head * fix typo * fix code * fix * fix typo * add detector * delete dectector * fix test_heads * add coder test * fix head * fix bugs in smoke head * refine kitti_mono3d data config * remove cuda is available * fix docstring * fix unitest * fix typo * refine * add affine aug * fix rebase typo * fix docs * change cam_intrinsics to cam2imgs * fix typos * fix lint * fix bugs * fix las typos * fix typos * add dosctrings for trans_mats Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: Xi Liu <75658786+xiliu8006@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> Co-authored-by: Wenhao Wu <79644370+wHao-Wu@users.noreply.github.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: 谢恩泽 <Johnny_ez@163.com> Co-authored-by: Robin Karlsson <34254153+robin-karlsson0@users.noreply.github.com> Co-authored-by: Danila Rukhovich <danrukh@gmail.com> * [Feature] add SMOKE augmentation method (#955) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * [Enhance] Change Groupfree3D config (#855) * All mods * PointSample * PointSample * [Doc] Add tutorials/data_pipeline Chinese version (#827) * [Doc] Add tutorials/data_pipeline Chinese version * refine doc * Use the absolute link * Use the absolute link Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add Chinese doc for `scannet_det.md` (#836) * Part * Complete * Fix comments * Fix comments * [Doc] Add Chinese doc for `waymo_det.md` (#859) * Add complete translation * Refinements * Fix comments * Fix a minor typo Co-authored-by: Tai-Wang <tab_wang@outlook.com> * Remove 2D annotations on Lyft (#867) * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [fix] fix typos (#872) * Fix 3 unworking configs (#882) * [Fix] Fix `index.rst` for Chinese docs (#873) * Fix index.rst for zh docs * Change switch language * [Fix] Centerpoint head nested list transpose (#879) * FIX Transpose nested lists without Numpy * Removed unused Numpy import * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Doc] Add nuscenes_det.md Chinese version (#854) * add nus chinese doc * add nuScenes Chinese doc * fix typo * fix typo * fix typo * fix typo * fix typo * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * [Enhance] Add benchmark regression script (#808) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo * fix typo * fix typo * support dgcnn * Change cam rot_3d_in_axis (#906) * [Doc] Add coord sys tutorial pic and change links to dev branch (#912) * Modify link branch and add pic * Fix pic * [Feature] add kitti AP40 evaluation metric (v1.0.0.dev0) (#927) * Add citation (#901) * [Feature] Add python3.9 in CI (#900) * Add python3.0 in CI * Add python3.0 in CI * Bump to v0.17.0 (#898) * Update README.md * Update README_zh-CN.md * Update version.py * Update getting_started.md * Update getting_started.md * Update changelog.md * Remove "recent" in the news * Remove "recent" in the news * Fix comments * [Docs] Fix the version of sphinx (#902) * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * add AP40 * add unitest * add unitest * seperate AP11 and AP40 * fix some typos Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Feature] add smoke backbone neck (#939) * add smoke detecotor and it's backbone and neck * typo fix * fix typo * add docstring * fix typo * fix comments * fix comments * fix comments * fix typo * fix typo * fix * fix typo * fix docstring * refine feature * fix typo * use Basemodule in Neck * [Refactor] Refactor the transformation from image to camera coordinates (#938) * Refactor points_img2cam * Refine docstring * Support array converter and add unit tests * [Feature] FCOS3D BBox Coder (#940) * FCOS3D BBox Coder * Add unit tests * Change the value from long to float/double * Rename bbox_out as bbox * Add comments to forward returns * [Feature] PGD BBox Coder (#948) * Support PGD BBox Coder * Refine docstring * add smoke augmentation method * add docs * fix docstrings * fix typos * change point name * fix typos * fix typos * fix typos Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: Xi Liu <75658786+xiliu8006@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> Co-authored-by: Wenhao Wu <79644370+wHao-Wu@users.noreply.github.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: 谢恩泽 <Johnny_ez@163.com> Co-authored-by: Robin Karlsson <34254153+robin-karlsson0@users.noreply.github.com> Co-authored-by: Danila Rukhovich <danrukh@gmail.com> * [Feature] Add SMOKE detector and Configs (#975) * add smoke detector and cfgs * fix init * fix input_modality * [Feature] Add SMOKE benchmark (#988) * add metafile and readme * fix typos * refine some typos * fix some typos * add smoke in readme and model-zoo * change dlanet to dla * change dlanet to dla in metafile * Add code-spell pre-commit hook and fix typos (#995) * [Feature] Add mmdet3d2torchserve tool (#977) * add tochserve support * fully support torchserve * Delete the empty file * fix typos * add docstrings and doc * add config.properties * fix typos and dosctrings * change pipeline name Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add coords system Chinese doc and fix English doc (#1001) * add coords Chinese doc and fix en doc * fix typos * fix typos and change latex insert methods * fix image show error * change coords image link * fix typos * fix typos * fix some typos * fix typos * [Feature] Support PointRCNN backbone (#974) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * support pointrcnn backbone * add docstring * modify docstring * modify docstring * modify docstring * Update pointnet2_fp_neck.py * add code block * refine docstring & code * add unittest on fp_neck * refine unittest * refine unittest * refine unittest * refine unittest * refine unittest * fix docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: Xi Liu <75658786+xiliu8006@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> Co-authored-by: xiliu8006 <xiliu8006@gmail.com> * fix smoke config link (#999) * [Feature] Support PGD Head (#964) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * [Enhance] Change Groupfree3D config (#855) * All mods * PointSample * PointSample * [Doc] Add tutorials/data_pipeline Chinese version (#827) * [Doc] Add tutorials/data_pipeline Chinese version * refine doc * Use the absolute link * Use the absolute link Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add Chinese doc for `scannet_det.md` (#836) * Part * Complete * Fix comments * Fix comments * [Doc] Add Chinese doc for `waymo_det.md` (#859) * Add complete translation * Refinements * Fix comments * Fix a minor typo Co-authored-by: Tai-Wang <tab_wang@outlook.com> * Remove 2D annotations on Lyft (#867) * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [fix] fix typos (#872) * Fix 3 unworking configs (#882) * [Fix] Fix `index.rst` for Chinese docs (#873) * Fix index.rst for zh docs * Change switch language * [Fix] Centerpoint head nested list transpose (#879) * FIX Transpose nested lists without Numpy * Removed unused Numpy import * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Doc] Add nuscenes_det.md Chinese version (#854) * add nus chinese doc * add nuScenes Chinese doc * fix typo * fix typo * fix typo * fix typo * fix typo * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * [Enhance] Add benchmark regression script (#808) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo * fix typo * fix typo * support dgcnn * Change cam rot_3d_in_axis (#906) * [Doc] Add coord sys tutorial pic and change links to dev branch (#912) * Modify link branch and add pic * Fix pic * [Feature] add kitti AP40 evaluation metric (v1.0.0.dev0) (#927) * Add citation (#901) * [Feature] Add python3.9 in CI (#900) * Add python3.0 in CI * Add python3.0 in CI * Bump to v0.17.0 (#898) * Update README.md * Update README_zh-CN.md * Update version.py * Update getting_started.md * Update getting_started.md * Update changelog.md * Remove "recent" in the news * Remove "recent" in the news * Fix comments * [Docs] Fix the version of sphinx (#902) * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * add AP40 * add unitest * add unitest * seperate AP11 and AP40 * fix some typos Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Feature] add smoke backbone neck (#939) * add smoke detecotor and it's backbone and neck * typo fix * fix typo * add docstring * fix typo * fix comments * fix comments * fix comments * fix typo * fix typo * fix * fix typo * fix docstring * refine feature * fix typo * use Basemodule in Neck * [Refactor] Refactor the transformation from image to camera coordinates (#938) * Refactor points_img2cam * Refine docstring * Support array converter and add unit tests * [Feature] FCOS3D BBox Coder (#940) * FCOS3D BBox Coder * Add unit tests * Change the value from long to float/double * Rename bbox_out as bbox * Add comments to forward returns * Support PGD BBox Coder * Refine docstring * Add uncertain l1 loss and its unit tests * [Feature] PGD BBox Coder (#948) * Support PGD BBox Coder * Refine docstring * PGD Head initialized * Refactor init methods, fix legacy variable names * [Feature] Support Uncertain L1 Loss (#950) * Add uncertain l1 loss and its unit tests * Remove mmcv.jit and refine docstrings * [Fix] Fix visualization in KITTI dataset (#956) * fix bug to support kitti vis * fix * Refine variable names and docstrings * Add unit tests and fix some minor bugs * Refine assertion messages * Fix typo in the docs_zh-CN * Use Pretrain init and remove unused init_cfg in FCOS3D * Fix the comments for the input_modality in the dataset config * Fix minor bugs in pgd_bbox_coder and incorrect setting for uncertain loss, use original init * Add explanations for code_weights * Adjust the unit test for pgd bbox coder * Remove unused codes * Add mono3d metric into the gather_models and fix bugs * Involve the value assignment of loss_dict into the computing procedure * Fix incorrect loss_depth Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: Xi Liu <75658786+xiliu8006@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> Co-authored-by: Wenhao Wu <79644370+wHao-Wu@users.noreply.github.com> Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: 谢恩泽 <Johnny_ez@163.com> Co-authored-by: Robin Karlsson <34254153+robin-karlsson0@users.noreply.github.com> Co-authored-by: Danila Rukhovich <danrukh@gmail.com> Co-authored-by: ChaimZhu <zhuchenming@pjlab.org.cn> * [Feature] PGD Benchmark on KITTI (#1014) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * [Enhance] Change Groupfree3D config (#855) * All mods * PointSample * PointSample * [Doc] Add tutorials/data_pipeline Chinese version (#827) * [Doc] Add tutorials/data_pipeline Chinese version * refine doc * Use the absolute link * Use the absolute link Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add Chinese doc for `scannet_det.md` (#836) * Part * Complete * Fix comments * Fix comments * [Doc] Add Chinese doc for `waymo_det.md` (#859) * Add complete translation * Refinements * Fix comments * Fix a minor typo Co-authored-by: Tai-Wang <tab_wang@outlook.com> * Remove 2D annotations on Lyft (#867) * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [fix] fix typos (#872) * Fix 3 unworking configs (#882) * [Fix] Fix `index.rst` for Chinese docs (#873) * Fix index.rst for zh docs * Change switch language * [Fix] Centerpoint head nested list transpose (#879) * FIX Transpose nested lists without Numpy * Removed unused Numpy import * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Doc] Add nuscenes_det.md Chinese version (#854) * add nus chinese doc * add nuScenes Chinese doc * fix typo * fix typo * fix typo * fix typo * fix typo * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * [Enhance] Add benchmark regression script (#808) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo * fix typo * fix typo * support dgcnn * Change cam rot_3d_in_axis (#906) * [Doc] Add coord sys tutorial pic and change links to dev branch (#912) * Modify link branch and add pic * Fix pic * [Feature] add kitti AP40 evaluation metric (v1.0.0.dev0) (#927) * Add citation (#901) * [Feature] Add python3.9 in CI (#900) * Add python3.0 in CI * Add python3.0 in CI * Bump to v0.17.0 (#898) * Update README.md * Update README_zh-CN.md * Update version.py * Update getting_started.md * Update getting_started.md * Update changelog.md * Remove "recent" in the news * Remove "recent" in the news * Fix comments * [Docs] Fix the version of sphinx (#902) * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * Fix sphinx version * add AP40 * add unitest * add unitest * seperate AP11 and AP40 * fix some typos Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Feature] add smoke backbone neck (#939) * add smoke detecotor and it's backbone and neck * typo fix * fix typo * add docstring * fix typo * fix comments * fix comments * fix comments * fix typo * fix typo * fix * fix typo * fix docstring * refine feature * fix typo * use Basemodule in Neck * [Refactor] Refactor the transformation from image to camera coordinates (#938) * Refactor points_img2cam * Refine docstring * Support array converter and add unit tests * [Feature] FCOS3D BBox Coder (#940) * FCOS3D BBox Coder * Add unit tests * Change the value from long to float/double * Rename bbox_out as bbox * Add comments to forward returns * Support PGD BBox Coder * Refine docstring * Add uncertain l1 loss and its unit tests * [Feature] PGD BBox Coder (#948) * Support PGD BBox Coder * Refine docstring * PGD Head initialized * Refactor init methods, fix legacy variable names * [Feature] Support Uncertain L1 Loss (#950) * Add uncertain l1 loss and its unit tests * Remove mmcv.jit and refine docstrings * [Fix] Fix visualization in KITTI dataset (#956) * fix bug to support kitti vis * fix * Refine variable names and docstrings * Add unit tests and fix some minor bugs * Refine assertion messages * Fix typo in the docs_zh-CN * Use Pretrain init and remove unused init_cfg in FCOS3D * Fix the comments for the input_modality in the dataset config * Fix minor bugs in pgd_bbox_coder and incorrect setting for uncertain loss, use original init * Add explanations for code_weights * Add PGD README * Adjust the unit test for pgd bbox coder * Remove unused codes * Add mono3d metric into the gather_models and fix bugs * Update README.md * Update links * Update links * Involve the value assignment of loss_dict into the computing procedure * Fix incorrect loss_depth * Update README.md * Update README_zh-CN.md * Update PGD in the model_zoo.md * Update PGD in the model_zoo.md * Update metafiles Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: Xi Liu <75658786+xiliu8006@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> Co-authored-by: Wenhao Wu <79644370+wHao-Wu@users.noreply.github.com> Co-authored-by: dingchang <hudingchang.vendor@sensetime.com> Co-authored-by: 谢恩泽 <Johnny_ez@163.com> Co-authored-by: Robin Karlsson <34254153+robin-karlsson0@users.noreply.github.com> Co-authored-by: Danila Rukhovich <danrukh@gmail.com> Co-authored-by: ChaimZhu <zhuchenming@pjlab.org.cn> * [Doc] Update compat doc on coord sys refactoring (#968) * Brief summary * v1 draft * Fix comments * [Fix] Support inference smoke on Nuscenes and fix bugs (#1029) * fix collect3d meta keys * support inference smoke on nuscenes and fix some bugs * fix some bugs * fix visualization bugs in mono3d (#1066) * support only save output meshlab results while evaluating mAP (#1073) * [Enchance] Set a random seed when the user does not set a seed. (#1072) * [Feature] Support PGD nuScenes benchmark (#1065) * Support PGD nuscenes benchmark * Update nuScenes benchmark * Update metafile.yml * Update model_zoo.md * Update model_zoo.md * Add 2x schedule configs * Update 2x schedule config links * Update 2x schedule meta infos * [Enhance] Update numba version (#1043) * update numba * update * fix numba version in doc * fix typo * fix_typos * [Fix] Remove `fp16` folder in `configs` (#1074) * remove fp16 folders * fix metafile.yml * fix typos * fix model_zoo.md * unify collections * fix typos * fix_typos * [Fix] Fix PointPillars FLOPs calculation error (#1075) * fix pointpillars flops calculation error * fix * [Enhance] Add abstract and a figure to provide a brief for each algorithm. (#1086) * add abstract for methods * fix typos * Add PGD to Docker image (#1070) Copy the repository in the Dockerfile instead of cloning. This makes it possible to work with latest modifications, such as PGD in the dev branch, and also network bandwith does not get wasted. * [Feature] Support PointRCNN RPN and RCNN module (#1022) * rebase & resubmit * rename config & model * fix unittest * resolve comments & add docstring for class_agnostic_nms * refine loss calculation & remove find_unused_parameters * resolve typo & add docstring * resolve comments * [Feature] Add MonoFlex data augmentation label generation function (#1026) * add pipeline * fix typos * fix unittest * fix_typos * fix_typos * change pipeline to func * refactor * refine gen indices * add device to gen indices * change tunc_objs_handle to handle_objs * fix comments * add numpy and tensor time comparision * fix * add numba accelerate * fix format * remove unnecssary func * update edge_indices docstrings * fix some comments * [Feature] Support PointRCNN KITTI benchmark (#1109) * [Feature] Support PointRCNN KITTI benchmark * update link to model&log * Remove redundant PointRCNN README.md (#1125) * Fix the path of coordinate system tutorials * [Fix] Fix SUN RGB-D generation (#1120) * [Feature] Add MonoFlex Loss (#1114) * add monoflex loss * add unnitest * add assert error * fix typos * remove monoflex_iou_loss * remove monoflex_iou_loss unittest * [Enhance] Add local yaw (alpha) property (#1130) * add local yaw property * fix typos * delete local yaw in base * change local yaw to alpha * change local yaw to alpha * update docstring * update docstring again * [Enhance] Support augmented training data and ground truth visualization (#1092) * support dataset augmentation visualization * unify two kinds of dataset visualization * update docs about visualization * fix typos * change description in doc * [Feature] Add MonoFlex Coder (#1115) * add monoflex coder * fix coder * fix comments * change variable name * fix typos * put encode alpha in coder * change variable name * change alpha to local yaw * fix roty to yaw * fix annotations * fix comments * fix return elements * fix comments * fix comments * Fix potential undefined variables in PGD when computing loss for cases with none positive predictions (#1157) * lock numba version (#1159) * [Feature] Add MonoFlex Head (#1044) * [Enhance] Support CPU Training (#1196) * add cpu support * fix * fix comments * [Fix] Update README format (#1195) * update readme * update readme * [Enhance] add README and copyright pre-commmit (#1134) * add precommmit to check readme * update * fix typos * [Enhance] add plane info in generated pkl file (#1162) * add plane info in pkl file * add_plane_info * fix comments * change to mmcv api * change road plane link * [Enhance] Accelerate training (#1168) * accelerate training * use multi_processes * get args from config * reorganize env setup * import from mmdet * import from mmdet * [Enhancement] Upgrade isort in pre-commit hook (#1217) * update isort * update setuptools * fix * pre-commit run --all-files * [Feature] add monoflex benchmark (#1183) * add monoflex benchmark * fix comments * fix comments * update descriptioin * [Fix] Fix a potential overflow bug when post_max_size is set too large in the circle_nms (#1225) * [Fix] overflow bug * [Fix] typo Co-authored-by: zeyuzeng <zeyuzeng@autox.ai> Co-authored-by: zeyu-hello <zeyuzeng@hotmail.com> * Fix PointRCNN bugs (#1224) * [Fix] Recheck import sorting (#1242) * fix missed mmcv * reinstall pre-commit and recheck * [Enhance] upgrade PointPillars performace on dev branch (#1166) * upgrade PointPillars performace on dev branch * update DynamicPillarFeatureNet * fix comments * change to AlignedAnchor3DRangeGenerator * change to AlignedAnchor3DRangeGenerator * fix * replace with AlignedAnchorGenerator * fix lint * update using isort * Bump to v1.0.0.rc0 (#928) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import (#839) * [Enhance] refactor iou_neg_piecewise_sampler.py (#842) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * refactor iou_neg_piecewise_sampler.py * add docstring * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Feature] Add roipooling cuda ops (#843) * [Refactor] Main code modification for coordinate system refactor (#677) * [Enhance] Add script for data update (#774) * Fixed wrong config paths and fixed a bug in test * Fixed metafile * Coord sys refactor (main code) * Update test_waymo_dataset.py * Manually resolve conflict * Removed unused lines and fixed imports * remove coord2box and box2coord * update dir_limit_offset * Some minor improvements * Removed some \s in comments * Revert a change * Change Box3DMode to Coord3DMode where points are converted * Fix points_in_bbox function * Fix Imvoxelnet config * Revert adding a line * Fix rotation bug when batch size is 0 * Keep sign of dir_scores as before * Fix several comments * Add a comment * Fix docstring * Add data update scripts * Fix comments * fix import * add roipooling cuda ops * add roi extractor * add test_roi_extractor unittest * Modify setup.py to install roipooling ops * modify docstring * remove enlarge bbox in roipoint pooling * add_roipooling_ops * modify docstring Co-authored-by: Yezhen Cong <52420115+THU17cyz@users.noreply.github.com> Co-authored-by: THU17cyz <congyezhen71@hotmail.com> * [Refactor] Refactor code structure and docstrings (#803) * refactor points_in_boxes * Merge same functions of three boxes * More docstring fixes and unify x/y/z size * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Add "optional" and fix "Default" * Remove None in function param type * Fix unittest * Add comments for NMS functions * Merge methods of Points * Add unittest * Add optional and default value * Fix box conversion and add unittest * Fix comments * Add unit test * Indent * Fix CI * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Remove useless \\ * Add unit test for box bev * More unit tests and refine docstrings in box_np_ops * Fix comment * Add deprecation warning * [Feature] PointXYZWHLRBBoxCoder (#856) * support PointBasedBoxCoder * fix unittest bug * support unittest in gpu * support unittest in gpu * modified docstring * add args * add args * [Enhance] Change Groupfree3D config (#855) * All mods * PointSample * PointSample * [Doc] Add tutorials/data_pipeline Chinese version (#827) * [Doc] Add tutorials/data_pipeline Chinese version * refine doc * Use the absolute link * Use the absolute link Co-authored-by: Tai-Wang <tab_wang@outlook.com> * [Doc] Add Chinese doc for `scannet_det.md` (#836) * Part * Complete * Fix comments * Fix comments * [Doc] Add Chinese doc for `waymo_det.md` (#859) * Add complete translation * Refinements * Fix comments * Fix a minor typo Co-authored-by: Tai-Wang <tab_wang@outlook.com> * Remove 2D annotations on Lyft (#867) * Add header for files (#869) * Add header for files * Add header for files * Add header for files * Add header for files * [fix] fix typos (#872) * Fix 3 unworking configs (#882) * [Fix] Fix `index.rst` for Chinese docs (#873) * Fix index.rst for zh docs * Change switch language * [Fix] Centerpoint head nested list transpose (#879) * FIX Transpose nested lists without Numpy * Removed unused Numpy import * [Enhance] Update PointFusion (#791) * update point fusion * remove LIDAR hardcode * move get_proj_mat_by_coord_type to utils * fix lint * remove todo * fix lint * [Doc] Add nuscenes_det.md Chinese version (#854) * add nus chinese doc * add nuScenes Chinese doc * fix typo * fix typo * fix typo * fix typo * fix typo * [Fix] Fix RegNet pretrained weight loading (#889) * Fix regnet pretrained weight loading * Remove unused file * Fix centerpoint tta (#892) * [Enhance] Add benchmark regression script (#808) * Initial commit * [Feature] Support DGCNN (v1.0.0.dev0) (#896) * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * support dgcnn * fix typo * fix typo * fix typo * del gf&fa registry (wo reuse pointnet module) * fix typo * add benchmark and add copyright header (for DGCNN only) * fix typo * fix typo * fix typo *…

ZwwWayne reviewed Nov 3, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

ZwwWayne reviewed Nov 3, 2021

View reviewed changes

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py Outdated Show resolved Hide resolved

ZwwWayne reviewed Nov 3, 2021

View reviewed changes

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py Outdated Show resolved Hide resolved

ZwwWayne reviewed Nov 3, 2021

View reviewed changes

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py Outdated Show resolved Hide resolved

ZCMax reviewed Nov 10, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

ZCMax reviewed Nov 10, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

Tai-Wang reviewed Nov 10, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

Tai-Wang requested changes Nov 10, 2021

View reviewed changes

wHao-Wu requested a review from Tai-Wang November 22, 2021 09:51

ZwwWayne requested review from ZCMax and ZwwWayne November 22, 2021 11:50

Tai-Wang reviewed Nov 23, 2021

View reviewed changes

mmdet3d/models/dense_heads/ssd_3d_head.py Outdated Show resolved Hide resolved

Tai-Wang reviewed Nov 23, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

rebase & resubmit

46f9654

wHao-Wu force-pushed the PointRCNN_RPN_RCNN branch from 24d4124 to 46f9654 Compare December 1, 2021 13:33

rename config & model

9b475a8

ZwwWayne approved these changes Dec 1, 2021

View reviewed changes

fix unittest

abc5f71

ZCMax approved these changes Dec 2, 2021

View reviewed changes

resolve comments & add docstring for class_agnostic_nms

2e5f90d

ZwwWayne reviewed Dec 7, 2021

View reviewed changes

mmdet3d/models/roi_heads/bbox_heads/point_rcnn_bbox_head.py Outdated Show resolved Hide resolved

refine loss calculation & remove find_unused_parameters

09134a2

Tai-Wang reviewed Dec 8, 2021

View reviewed changes

configs/_base_/models/point_rcnn.py Outdated Show resolved Hide resolved

Tai-Wang reviewed Dec 8, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Show resolved Hide resolved

Tai-Wang reviewed Dec 8, 2021

View reviewed changes

mmdet3d/models/detectors/point_rcnn.py Outdated Show resolved Hide resolved

resolve typo & add docstring

bbd545c

Tai-Wang reviewed Dec 8, 2021

View reviewed changes

mmdet3d/models/dense_heads/point_rpn_head.py Outdated Show resolved Hide resolved

resolve comments

6781b91

Tai-Wang approved these changes Dec 8, 2021

View reviewed changes

ZwwWayne merged commit 0f8181f into open-mmlab:v1.0.0.dev0 Dec 8, 2021

Machine97 mentioned this pull request Dec 16, 2021

Error on training PointRCNN #1118

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Feature] Support PointRCNN RPN and RCNN module #1022

[Feature] Support PointRCNN RPN and RCNN module #1022

wHao-Wu commented Nov 1, 2021

CLAassistant commented Dec 1, 2021 •

edited

Loading

Machine97 commented Dec 2, 2021

wHao-Wu commented Dec 3, 2021

ZwwWayne commented Dec 7, 2021

Machine97 commented Dec 14, 2021

Machine97 commented Dec 14, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

wHao-Wu commented Dec 16, 2021

wHao-Wu commented Dec 16, 2021 •

edited

Loading

Machine97 commented Dec 16, 2021

wHao-Wu commented Dec 17, 2021

Machine97 commented Dec 17, 2021

Machine97 commented Dec 18, 2021

wHao-Wu commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 23, 2021

[Feature] Support PointRCNN RPN and RCNN module #1022

[Feature] Support PointRCNN RPN and RCNN module #1022

Conversation

wHao-Wu commented Nov 1, 2021

Motivation

Modification

BC-breaking (Optional)

Use cases (Optional)

Checklist

CLAassistant commented Dec 1, 2021 • edited Loading

Machine97 commented Dec 2, 2021

wHao-Wu commented Dec 3, 2021

ZwwWayne commented Dec 7, 2021

Machine97 commented Dec 14, 2021

Machine97 commented Dec 14, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

Machine97 commented Dec 15, 2021

wHao-Wu commented Dec 16, 2021

wHao-Wu commented Dec 16, 2021 • edited Loading

Machine97 commented Dec 16, 2021

wHao-Wu commented Dec 17, 2021

Machine97 commented Dec 17, 2021

Machine97 commented Dec 18, 2021

wHao-Wu commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 20, 2021

Machine97 commented Dec 23, 2021

CLAassistant commented Dec 1, 2021 •

edited

Loading

wHao-Wu commented Dec 16, 2021 •

edited

Loading