- FCN: Fully convolutional networks for semantic segmentation
- fully-connected CRF: Efficient inference in fully connected crfs with gaussian edge potentials
- FCN + fully-connected CRF: Semantic image segmentation with deep convolutional nets and fully connected crfs
- PSPnet: Pyramid Scene Parsing Network
| Year | Paper | Descriptor |
|---|---|---|
| 2016 TOG | AutoHair: Fully Automatic Hair Modeling from A Single Image | FCN + CRF |
| 2016 | Hair Segmentation Using Heuristically-Trained Neural Networks | ... |
| 2017 | Hair segmentation using adaptive threshold from edge and branch length measures | ... |
| 2017 CVPR | Deep Image Matting | |
| 2017 ICME | Automatic skin and hair masking using fully convolutional networks | FCN + CRF |
| 2018 Image and Vision Computing |
Hair detection, segmentation, and hairstyle classification in the wild | ... |
| 2018 ECCV | HairNet: Single-View Hair Reconstruction using Convolutional Neural Networks | PSPnet |
| 2018 CRV | Real-Time Deep Hair Matting on Mobile Devices | ... |