diff --git a/README.md b/README.md index f83bc21..03f5493 100644 --- a/README.md +++ b/README.md @@ -26,52 +26,28 @@ EasyPortrait dataset size is about **26GB**, and it contains **20 000** RGB imag ... ``` -## Training, Evaluation and Testing on EasyPortrait - ->The code is based on [MMSegmentation](https://github.com/open-mmlab/mmsegmentation) with 0.30.0 version. - -Models were trained and evaluated on 8 NVIDIA V100 GPUs with CUDA 11.2 - -For installation process follow the instructions [here](https://github.com/open-mmlab/mmsegmentation/blob/v0.30.0/docs/en/get_started.md#installation) and use the **requirements.txt** file in our repository. - -### Training -For single GPU mode: -```console -python ./pipelines/tools/train.py ./pipelines/local_configs/easy_portrait_experiments//.py --gpu-id -``` - -For distributed training mode: -```console -./pipelines/tools/dist_train.sh ./pipelines/local_configs/easy_portrait_experiments//.py -``` -### Evaluation -For single GPU mode: -```console -python ./pipelines/tools/test.py --gpu-id --eval mIoU -``` - -For distributed evaluation mode: -```console -./pipelines/tools/dist_test.sh --eval mIoU -``` -### Run demo -```console -python ./pipelines/demo/image_demo.py --palette=easy_portrait --out-file= -``` - - - ## Models We provide some pre-trained models as the baseline for portrait segmentation and face parsing. We use mean Intersection over Union (mIoU) as the main metric. | Model Name | Parameters (M) | Input shape | mIOU | |------------------------------------------------|----------------|-------------|-----------| | [LR-ASPP + MobileNet-V3](https://sc.link/gBo6) | 1.14 | 1024 × 1024 | 73.13 | +| [FCN + MobileNet-V2](https://sc.link/ErPv) | 9.71 | 384 × 384 | 74.3 | +| [FCN + MobileNet-V2](https://sc.link/vKjm) | 9.71 | 512 × 512 | 77.01 | | [FCN + MobileNet-V2](https://sc.link/9xZ3) | 9.71 | 1024 × 1024 | 81.23 | -| [FPN + ResNet-50](https://sc.link/6r19) | 28.5 | 512 × 512 | **83.13** | +| [FPN + ResNet-50](https://sc.link/6r19) | 28.5 | 512 × 512 | 83.13 | +| [FPN + ResNet-50](https://sc.link/Gy97) | 28.5 | 1024 × 1024 | **85.97** | +| [BiSeNet-V2](https://sc.link/ryYE) | 14.79 | 512 × 512 | 77.93 | | [BiSeNet-V2](https://sc.link/8wZo) | 14.79 | 1024 × 1024 | 69.13 | +| [SegFormer-B0](https://sc.link/wMkR) | 3.72 | 384 × 384 | 79.82 | | [SegFormer-B0](https://sc.link/0lZX) | 3.72 | 1024 × 1024 | 73.41 | +| [SegFormer-B2](https://sc.link/AjmO) | 24.73 | 384 × 384 | 81.59 | +| [SegFormer-B2](https://sc.link/zVnY) | 24.73 | 512 × 512 | 83.03 | | [SegFormer-B2](https://sc.link/7vZA) | 24.73 | 1024 × 1024 | 76.19 | +| [SegFormer-B5](https://sc.link/yQm7) | 81.97 | 384 × 384 | 81.66 | +| [SegFormer-B5](https://sc.link/xOl9) | 81.97 | 1024 × 1024 | 85.80 | +| [SegNeXt + MSCAN-T](https://sc.link/Dp0x) | 4.23 | 384 × 384 | 75.01 | +| [SegNeXt + MSCAN-T](https://sc.link/BlnX) | 4.23 | 512 × 512 | 78.59 | ## Annotations @@ -109,6 +85,41 @@ where: ## Images ![easyportrait](images/data.jpg) + +## Training, Evaluation and Testing on EasyPortrait + +>The code is based on [MMSegmentation](https://github.com/open-mmlab/mmsegmentation) with 0.30.0 version. + +Models were trained and evaluated on 8 NVIDIA V100 GPUs with CUDA 11.2. + +For installation process follow the instructions [here](https://github.com/open-mmlab/mmsegmentation/blob/v0.30.0/docs/en/get_started.md#installation) and use the **requirements.txt** file in our repository. + +### Training +For single GPU mode: +```console +python ./pipelines/tools/train.py ./pipelines/local_configs/easy_portrait_experiments//.py --gpu-id +``` + +For distributed training mode: +```console +./pipelines/tools/dist_train.sh ./pipelines/local_configs/easy_portrait_experiments//.py +``` + +### Evaluation +For single GPU mode: +```console +python ./pipelines/tools/test.py --gpu-id --eval mIoU +``` + +For distributed evaluation mode: +```console +./pipelines/tools/dist_test.sh --eval mIoU +``` +### Run demo +```console +python ./pipelines/demo/image_demo.py --palette=easy_portrait --out-file= +``` + ## Authors and Credits - [Alexander Kapitanov](https://www.linkedin.com/in/hukenovs) - [Karina Kvanchiani](https://www.linkedin.com/in/kvanchiani)