Research map / Image generation / GANs
Shadow Detection With Conditional Generative Adversarial Networks
Vu Nguyen, Tomas F. Yago Vicente, Maozheng Zhao, Minh Hoai, Dimitris Samaras
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add sensitivity parameter and weighted data loss to cGAN (scGAN vs cGAN) (ablation) | cGAN · SBU test | BER: 13.6 -> 9.1 | Improved |
| add sensitivity parameter and weighted data loss to cGAN (ablation) | cGAN · SBU test | shadow error rate: 20.5 -> 7.8 | Improved |
| add sensitivity parameter and weighted data loss to cGAN (ablation) | cGAN · SBU test | non-shadow error rate: 6.9 -> 10.4 | Hurt |
| raise test-time sensitivity w from 0.3 to 0.9 (ablation) | scGAN · SBU test | shadow error rate: 15.4 -> 4.7 | Improved |
| raise test-time sensitivity w from 0.3 to 0.9 (ablation) | scGAN · SBU test | non-shadow error rate: 5.8 -> 16.5 | Hurt |
| choose w = 0.7 instead of 0.3 (ablation) | scGAN · SBU test | BER: 10.6 -> 9.0 | Improved |
| choose w = 0.9 instead of 0.7 (ablation) | scGAN · SBU test | BER: 9.0 -> 10.6 | Hurt |
| use scGAN instead of stackedCNN | StackedCNN · SBU test | BER: 11.0 -> 9.1 | Improved |
| use scGAN instead of stackedCNN, both trained on UCF | StackedCNN · UCF test | BER: 11.6 -> 10.9 | Improved |
| use scGAN instead of stackedCNN, both trained on SBU | StackedCNN · UCF test (cross-dataset) | BER: 13.0 -> 11.5 | Improved |
Closest papers by meaning
- Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removal (CVPR 2018 · similarity 0.94)
- ARGAN: Attentive Recurrent Generative Adversarial Network for Shadow Detection and Removal (ICCV 2019 · similarity 0.90)
- RoCGAN: Robust Conditional GAN (IJCV 2020 · similarity 0.89)
- A+D Net: Training a Shadow Detector with Adversarial Shadow Attenuation (ECCV 2018 · similarity 0.89)
- Mask-ShadowGAN: Learning to Remove Shadows From Unpaired Data (ICCV 2019 · similarity 0.89)
- Taming the Tail in Class-Conditional GANs: Knowledge Sharing via Unconditional Training at Lower Resolutions (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
