Segmenter: Transformer for Semantic Segmentation by Robin Strudel*, Ricardo Garcia*, Ivan Laptev and Cordelia Schmid, ICCV 2021.

*Equal Contribution

Define os environment variables pointing to your checkpoint and dataset directory, put in your .bashrc:

export DATASET=/path/to/dataset/dir

Install PyTorch 1.9 then pip install . at the root of this repository.

To download ADE20K, use the following command:

python -m segm.scripts.prepare_ade20k $DATASET

We release models with a Vision Transformer backbone initialized from the improved ViT models.

Segmenter models with ViT backbone:

Segmenter models with DeiT backbone:

Download one checkpoint with its configuration in a common folder, for example seg_tiny_mask.

You can generate segmentation maps from your own data with:

python -m segm.inference --model-path seg_tiny_mask/checkpoint.pth -i images/ -o segmaps/ 

To evaluate on ADE20K, run the command:

# single-scale evaluation:
python -m segm.eval.miou seg_tiny_mask/checkpoint.pth ade20k --singlescale
# multi-scale evaluation:
python -m segm.eval.miou seg_tiny_mask/checkpoint.pth ade20k --multiscale

Train Seg-T-Mask/16 on ADE20K on a single GPU:

python -m segm.train --log-dir seg_tiny_mask --dataset ade20k \
  --backbone vit_tiny_patch16_384 --decoder mask_transformer

To train Seg-B-Mask/16, simply set vit_base_patch16_384 as backbone and launch the above command using a minimum of 4 V100 GPUs (~12 minutes per epoch) and up to 8 V100 GPUs (~7 minutes per epoch). The code uses SLURM environment variables.

To plot the logs of your experiments, you can use

python -m segm.utils.logs logs.yml

with logs.yml located in utils/ with the path to your experiments logs:

root: /path/to/checkpoints/
logs:
  seg-t: seg_tiny_mask/log.txt
  seg-b: seg_base_mask/log.txt

To visualize the attention maps for Seg-T-Mask/16 encoder layer 0 and patch (0, 21), you can use:

python -m segm.scripts.show_attn_map seg_tiny_mask/checkpoint.pth \ 
images/im0.jpg output_dir/ --layer-id 0 --x-patch 0 --y-patch 21 --enc

Different options are provided to select the generated attention maps:

• --enc or --dec: Select encoder or decoder attention maps respectively.
• --patch or --cls: --patch generates attention maps for the patch with coordinates (x_patch, y_patch). --cls combined with --enc generates attention maps for the CLS token of the encoder. --cls combined with --dec generates maps for each class embedding of the decoder.
• --x-patch and --y-patch: Coordinates of the patch to draw attention maps from. This flag is ignored when --cls is used.
• --layer-id: Select the layer for which the attention maps are generated.

For example, to generate attention maps for the decoder class embeddings, you can use:

python -m segm.scripts.show_attn_map seg_tiny_mask/checkpoint.pth \
images/im0.jpg output_dir/ --layer-id 0 --dec --cls

Attention maps for patch (0, 21) in Seg-L-Mask/16 encoder layers 1, 4, 8, 12 and 16:

Attention maps for the class embeddings in Seg-L-Mask/16 decoder layer 0:

Zero shot video segmentation on DAVIS video dataset with Seg-B-Mask/16 model trained on ADE20K.

@article{strudel2021,
  title={Segmenter: Transformer for Semantic Segmentation},
  author={Strudel, Robin and Garcia, Ricardo and Laptev, Ivan and Schmid, Cordelia},
  journal={arXiv preprint arXiv:2105.05633},
  year={2021}
}

The Vision Transformer code is based on timm library and the semantic segmentation training and evaluation pipeline is using mmsegmentation.