retinanet

Focal Loss for Dense Object Detection

Introduction

@inproceedings{lin2017focal,
  title={Focal loss for dense object detection},
  author={Lin, Tsung-Yi and Goyal, Priya and Girshick, Ross and He, Kaiming and Doll{\'a}r, Piotr},
  booktitle={Proceedings of the IEEE international conference on computer vision},
  year={2017}
}

Results and models

Backbone	Style	Lr schd	Mem (GB)	Inf time (fps)	box AP	Download
R-50-FPN	caffe	1x	3.5	18.6	36.3	model | log
R-50-FPN	pytorch	1x	3.8	19.0	36.5	model | log
R-50-FPN	pytorch	2x	-	-	37.4	model | log
R-101-FPN	caffe	1x	5.5	14.7	38.5	model | log
R-101-FPN	pytorch	1x	5.7	15.0	38.5	model | log
R-101-FPN	pytorch	2x	-	-	38.9	model | log
X-101-32x4d-FPN	pytorch	1x	7.0	12.1	39.9	model | log
X-101-32x4d-FPN	pytorch	2x	-	-	40.1	model | log
X-101-64x4d-FPN	pytorch	1x	10.0	8.7	41.0	model | log
X-101-64x4d-FPN	pytorch	2x	-	-	40.8	model | log