Research map / 3D vision / Depth and stereo
Learn to See the Unseen in Low-light Spike Streams
Liwen Hu, Yang Li, Mianzhi Liu, Guo Yijia, Shenghao Xie, Wenqiang Zu, gang ding, Tiejun Huang, Lei Ma
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 |
|---|---|---|---|
| use ETFI as ControlNet condition instead of a prior reconstruction network output (ablation) | Diff-SPK · SAT-SPK (synthetic) | PSNR: 15.605 -> 23.499 | Improved |
| use ETFI as condition (ablation) | Diff-SPK · OutdoorSPK (real) | MUSIQA: 50.895 -> 49.494 | Mixed |
| use tailored condition encoder instead of fine-tuned VAE encoder (ablation) | Diff-SPK · SAT-SPK (synthetic) | PSNR: 21.891 -> 23.499 | Improved |
| use tailored condition encoder instead of fine-tuned VAE encoder (ablation) | Diff-SPK · SAT-SPK (synthetic) | LPIPS: 0.353 -> 0.223 | Improved |
| add ETFI-based fusion module (ablation) | Diff-SPK · SAT-SPK (synthetic) | PSNR: 23.365 -> 23.499 | Improved |
| add ETFI-based fusion module (ablation) | Diff-SPK · IndoorSPK (real) | MANIQA: 0.408 -> 0.436 | Improved |
| use Diff-SPK instead of regression methods on LLR (brighter scenes) | Hu et al. · LLR (synthetic) | PSNR: 45.075 -> 37.898 | Hurt |
| use Diff-SPK instead of regression methods on LLR (brighter scenes) | Hu et al. · LLR (synthetic) | MUSIQA: 49.927 -> 52.780 | Improved |
| use Diff-SPK on darker synthetic SAT-SPK vs best baseline | TFI (best baseline) · SAT-SPK (synthetic) | PSNR: 21.909 -> 23.499 | Improved |
| use Diff-SPK across 40/60/80 Lux illumination | Hu et al. · IndoorSPK 80 Lux | MUSIQA: 45.980 -> 50.506 | Improved |
| increase classifier-free guidance scale (ablation) | Diff-SPK · LLR (synthetic) | MUSIQA / PSNR | Mixed |
Closest papers by meaning
- Rethinking High-speed Image Reconstruction Framework with Spike Camera (AAAI 2025 · similarity 0.91)
- Spatio-Temporal Interactive Learning for Efficient Image Reconstruction of Spiking Cameras (NeurIPS 2024 · similarity 0.90)
- SpikeNeRF: Learning Neural Radiance Fields from Continuous Spike Stream (CVPR 2024 · similarity 0.89)
- Finding Visual Saliency in Continuous Spike Stream (AAAI 2024 · similarity 0.88)
- Spk2SRImgNet: Super-Resolve Dynamic Scene from Spike Stream via Motion Aligned Collaborative Filtering (CVPR 2025 · similarity 0.88)
- SpikeReveal: Unlocking Temporal Sequences from Real Blurry Inputs with Spike Streams (NeurIPS 2024 · similarity 0.88)
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.
