Research map / Video / Video segmentation and tracking
Pol-CACTI: A System and dataset forHigh-Speed Polarized Video Compressive Imaging
Yunfeng Song, Yidong Luo, Ping Wang, xingjian jiang, Xin Yuan
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 PA-LBP module (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 30.26 -> 39.79 | Improved |
| add PA-LBP module (ablation) | Pol-CACTI network · Pol-CACTI dataset | AoP MAE (deg): 39.09 -> 9.33 | Improved |
| use PA-LBP instead of a deeper ResNet (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 32.37 -> 39.79 | Improved |
| add physics-constrained losses (ablation) | Pol-CACTI network · Pol-CACTI dataset | DoP PSNR (dB): 21.99 -> 30.64 | Improved |
| add physics-constrained losses (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 39.47 -> 39.79 | No clear effect |
| add physics-constrained losses (ablation) | Pol-CACTI network · Pol-CACTI dataset | AoP MAE (deg): 14.50 -> 9.33 | Improved |
| use joint reconstruction and demosaicing instead of serial cascade (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 15.78 -> 39.79 | Improved |
| use progressive staged training instead of joint training from scratch (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 33.16 -> 39.79 | Improved |
| add end-to-end fine-tuning stage after frozen backbone (ablation) | Pol-CACTI network · Pol-CACTI dataset | S0 PSNR (dB): 35.91 -> 39.79 | Improved |
| add end-to-end fine-tuning stage after frozen backbone (ablation) | Pol-CACTI network · Pol-CACTI dataset | AoP MAE (deg): 14.48 -> 9.33 | Improved |
| use joint network instead of Polanalyser (ablation) | reconstruction · Simulated test (Cr=8) | S0 PSNR (dB): 36.384 -> 42.122 | Improved |
| increase compression ratio from 8 to 48 (ablation) | Pol-CACTI network · Simulated test | S0 PSNR (dB): 42.122 -> 33.507 | Hurt |
Closest papers by meaning
- 340 FPS Reflection-free Video from Spikes Modulated by a Rapidly Rotating Polarizer (ECCV 2026 · similarity 0.87)
- 3One2: One-Step Regression plus One-Step Diffusion for One-Hot Modulation in Dual-Path Video Snapshot Compressive Imaging (AAAI 2026 · similarity 0.87)
- Stokes-Informed Diffusion for Robust Linear Polarization Estimation (ECCV 2026 · similarity 0.86)
- Toward a Unified Complementary Fusion Framework for Robust Polarimetric Imaging (TPAMI 2026 · similarity 0.86)
- PolarScale: A Physics-Grounded Benchmark for Radiometrically Consistent RGB-to-Stokes Estimation (NeurIPS 2026 · similarity 0.86)
- RGB-to-Polarization Estimation: A New Task and Benchmark Study (NeurIPS 2025 · similarity 0.85)
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.
