Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learning to See in the Dark with Events
Song Zhang, Yu Zhang, Zhe Jiang, Dongqing Zou, Jimmy Ren, Bin Zhou
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 noise channel (ablation) | Event-to-image domain adaptation · DVS-Dark simulated | PSNR: 20.69 -> 21.24 | Improved |
| add detail enhancing branch (ablation) | Event-to-image domain adaptation · DVS-Dark simulated | PSNR: 21.24 -> 24.91 | Improved |
| add ranking regularization (ablation) | Event-to-image domain adaptation · DVS-Dark simulated | PSNR: 24.91 -> 26.03 | Improved |
| add ranking regularization (ablation) | Event-to-image domain adaptation · DVS-Dark simulated | SSIM: 0.75 -> 0.77 | Improved |
| add detail recovery on top of shared-space reconstruction (SDA) | SDA · DVS-Dark simulated | PSNR: 24.93 -> 26.03 | Improved |
| add domain adaptation to Pix2Pix baseline | Pix2Pix · DVS-Dark simulated | PSNR: 22.65 -> 26.03 | Improved |
| fine-tune E2V on day-light images of DVS-Dark | E2V · DVS-Dark simulated | PSNR: 12.16 -> 17.15 | Improved |
| use ranking loss weight 5.0 instead of 1.0 (ablation) | Ranking regularization · DVS-Dark | visual quality | Hurt |
| use ranking loss weight 1.0 (ablation) | Ranking regularization · DVS-Dark | visual quality | Improved |
| use ours instead of Pix2Pix on real low-light events | Pix2Pix · DVS-Dark real low light | FID: 109.08 -> 104.21 | Improved |
| use ours instead of E2V on real low-light events | E2V · DVS-Dark real low light | FID: 210.42 -> 104.21 | Improved |
| train with ranking loss margin 0.5 (ablation) | Ranking regularization · DVS-Dark | PSNR | No clear effect |
Closest papers by meaning
- Zero-Shot Day-Night Domain Adaptation With a Physics Prior (ICCV 2021 · similarity 0.87)
- NER-Net+: Seeing Motion at Nighttime With an Event Camera (TPAMI 2025 · similarity 0.87)
- Seeing Motion at Nighttime with an Event Camera (CVPR 2024 · similarity 0.87)
- Boosting Object Detection with Zero-Shot Day-Night Domain Adaptation (CVPR 2024 · similarity 0.87)
- Similarity Min-Max: Zero-Shot Day-Night Domain Adaptation (ICCV 2023 · similarity 0.86)
- Towards Robust Event-guided Low-Light Image Enhancement: A Large-Scale Real-World Event-Image Dataset and Novel Approach (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.
