Explicit Visual Prompting for Low-Level Structure Segmentations

Weihuang Liu ¹   Xi Shen ²  Chi-Man Pun ^*,1  Xiaodong Cun ^*,2 

¹ University of Macau   ² Tencent AI Lab  

CVPR 2023

We consider the generic problem of detecting low-level structures in images, which includes segmenting the manipulated parts, identifying out-of-focus pixels, separating shadow regions, and detecting concealed objects. Whereas each such topic has been typically addressed with a domain-specific solution, we show that a unified approach performs well across all of them. We take inspiration from the widely-used pre-training and then prompt tuning protocols in NLP and propose a new visual prompting model, named Explicit Visual Prompting (EVP). Different from the previous visual prompting which is typically a dataset-level implicit embedding, our key insight is to enforce the tunable parameters focusing on the explicit visual content from each individual image, i.e., the features from frozen patch embeddings and the input's high-frequency components. The proposed EVP significantly outperforms other parameter-efficient tuning protocols under the same amount of tunable parameters (5.7% extra trainable parameters of each task). EVP also achieves state-of-the-art performances on diverse low-level structure segmentation tasks compared to task-specific solutions.

This code was implemented with Python 3.6 and PyTorch 1.8.1. You can install all the requirements via:

pip install -r requirements.txt

python demo.py --input [INPUT_PATH] --model [MODEL_PATH] --prompt [PROMPT_PATH] --resolution [HEIGHT],[WIDTH] --config [CONFIG_PATH]

[INPUT_PATH]: input image [PROMPT_PATH]: prompt checkpoint [MODEL_PATH]: backbone checkpoint [HEIGHT]: target height [WIDTH]: target width [CONFIG_PATH]: config file

1. Download the dataset and put it in ./load.
2. Download the pre-trained SegFormer backbone.
3. Training:

python train.py --config configs/train/segformer/train_segformer_evp_defocus.yaml 

4. Evaluation:

python test.py --config configs/test/test_defocus.yaml  --model mit_b4.pth --prompt ./save/_train_segformer_evp_defocus/prompt_epoch_last.pth

5. Visualization:

python demo.py --input defocus.png --model ./mit_b4.pth --prompt ./save/_train_segformer_evp_defocus/prompt_epoch_last.pth --resolution 320,320 --config configs/demo.yaml

python train.py --config [CONFIG_PATH]

python test.py --config [CONFIG_PATH] --model [MODEL_PATH] --prompt [PROMPT_PATH]

Please find the pre-trained models here.

• COD10K: https://github.com/DengPingFan/SINet/
• CAMO: https://drive.google.com/open?id=1h-OqZdwkuPhBvGcVAwmh0f1NGqlH_4B6
• CHAMELEON: https://www.polsl.pl/rau6/datasets/

• DUT: http:https://ice.dlut.edu.cn/ZhaoWenda/BTBCRLNet.html
• CUHK: http:https://www.cse.cuhk.edu.hk/leojia/projects/dblurdetect/

• CAISA: https://github.com/namtpham/casia2groundtruth
• IMD2020: http:https://staff.utia.cas.cz/novozada/db/

• ISTD: https://github.com/DeepInsight-PCALab/ST-CGAN
• SBU: https://www3.cs.stonybrook.edu/~cvl/projects/shadow_noisy_label/index.html

If you find our work useful in your research, please consider citing:

@inproceedings{liu2023evp,
  title={Explicit Visual Prompting for Low-Level Structure Segmentations},
  author={Liu, Weihuang and Shen, Xi and Pun, Chi-Man and Cun, Xiaodong},
  booktitle={CPVR},
  year={2023}
}

EVP code borrows heavily from LIIF, SETR and SegFormer. We thank the author for sharing their wonderful code.