[TCSVT2026] [iHEPNet] Exploring Information Entropy-driven Interaction and Hierarchical Edge Perception for Lightweight ORSI Salient Object Detection PDF|Homepage

Network Architecture

Requirements

python 3.8 + pytorch 1.13.1

Saliency maps

We provide saliency maps of our iHEPNet, lightweight methods (code: frem), and normal-size methods (code: frem) on the ORSSD, EORSSD, and ORSI-4199 datasets.

Training

We use data_aug.m for data augmentation.

Modify paths of datasets, then run train_iHEPNet.py.

Note: Our main model is under './model/GeleNet_models.py'. Our code is built on GeleNet. So in this code, GeleNet refers to our iHEPNet.

Pre-trained model and testing

1. We provide the pre-trained models in './models/'.

2. Modify paths of pre-trained models and datasets.

3. Run test_iHEPNet.py.

Evaluation Tool

You can use the evaluation tool (MATLAB version) to evaluate the above saliency maps.

ORSI-SOD_Summary

Citation

    @ARTICLE{Li_2026_iHEPNet,
              author={Yihua Tu and Wenqi Si and Gongyang Li and Chengjun Han and Yun Sui and Weisi Lin},
              title={Exploring Information Entropy-driven Interaction and Hierarchical Edge Perception for Lightweight {ORSI} Salient Object Detection}, 
              journal={IEEE Transactions on Circuits and Systems for Video Technology}, 
              year={2026},
              pages={1-13},
            }
            
            

If you encounter any problems with the code, want to report bugs, etc.

Please contact me at lllmiemie@163.com or ligongyang@shu.edu.cn.