ESCNet:Edge-Semantic Collaborative Network for Camouflaged Object Detection

This repository is the official implementation of ESCNet:Edge-Semantic Collaborative Network for Camouflaged Object Detection

Requirements

• python == 3.11
• cuda >= 12.4

To install requirements:

pip install -r requirements.txt

Dataset

COD (Camouflaged Object Detection) Dataset

Training

To train the model(s) in the paper, run this command:

torchrun --nproc_per_node=4 train.py --config config.yaml

Evaluation

To test models, change config.yaml for different datasets:

# inference preds on different model
python test.py --config config.yaml --pred_root preds
# Then calculate metrics
python eval.py --pred_root preds ----save_dir results

For ease of use, we create a eval.sh script and a use case in the form of a shell script eval.sh. You can edit the script to change the parameters you want to test.

bash run.sh
# for eval only
bash run.sh --notrain

Tips