We make the study to unify SOD and COD in a task-agnostic framework via a contrastive distillation paradigm, inspired by their agreeable nature of binary segmentation.

💥 News 💥

• [13.2.2025] Seamless-Detection has been accepted to ESWA 2025 !
• [16.12.2024] Paper is now available on arXiv !
• [04.9.2024] We have released the code and model checkpoints for Seamless-Detection !

Quick Start

• Pretrained backbone:MoCo-v2.

• Training

python train_compare.py \
cornet_compare \
--gpus=0 \
--save \
--found \
--vals=ECSSD

• Test

python test.py \
cornet_compare \
--weight="./weight/cornet_compare/resnet/base/cornet_compare_base_24.pth" \
--gpus=0 \
--save \
--vals=ECSSD,DUTS-TE,DUT-OMRON,PASCAL-S

Requirements

python 3.9

pytorch 1.11.0

tensorboardX 2.5

Weights and Results

Baidu | 提取码:sldt

Citation

If you find Seamless-Detection to be useful for your work, please consider citing our paper:

 @article{liu2025seamless,
   title={Seamless Detection: Unifying Salient Object Detection and Camouflaged Object Detection},
   author={Liu, Yi and Li, Chengxin and Dong, Xiaohui and Li, Lei and Zhang, Dingwen and Xu, Shoukun and Han, Jungong},
   journal={Expert Systems with Applications},
   pages={126912},
   year={2025},
   publisher={Elsevier}
 }