Research map / Video / Action recognition
Self-Conditioned Probabilistic Learning of Video Rescaling
Yuan Tian, Guo Lu, Xiongkuo Min, Zhaohui Che, Guangtao Zhai, Guodong Guo, Zhiyong Gao
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 temporal modeling by switching Dense2D to Dense2D-T (ablation) | SelfC-basic · Vid4-Y | PSNR-Y: 30.62 -> 31.29 | Improved |
| increase GMM mixture number from 1 to 5 (ablation) | SelfC-basicT · Vid4-Y | PSNR-Y: 31.29 -> 31.61 | Improved |
| enlarge backbone 4 times (2 to 8 Dense2D-T blocks) (ablation) | SelfC-small · Vid4-Y | PSNR-Y: 31.61 -> 32.30 | Improved |
| add self-conditioned STP-Net probability model to auto-encoder (ablation) | Auto-Enc · Vid4-Y | PSNR-Y: 28.91 -> 31.61 | Improved |
| replace Dense2D backbone with temporal Dense2D-T in IRN (ablation) | IRN · Vid4-Y | PSNR-Y: 30.68 -> 30.42 | Hurt |
| use SelfC instead of IRN for rescaling | SelfC-large · Vid4 | PSNR-Y: 31.29 -> 32.30 | Improved |
| use proposed control-variates gradient estimator instead of straight-through estimator for H.265 (ablation) | SelfC compression · video compression | PSNR: nearly 0.3dB | Improved |
| add SelfC rescaling before H.265 (default mode) | H.265 · UVG | BDBR-SSIM: -50.84 | Improved |
| add SelfC rescaling before H.265 (zerolatency mode) | H.265 · HEVC Class B | BDBR-SSIM: -18.04 | Improved |
| use learned SelfC downscaler instead of bicubic, no fine-tuning | TSM · Something V1 | Top-1 accuracy: 30.4 -> 32.7 | Improved |
| use learned downscaler instead of bicubic with action CNN fine-tuned | TSM · Something V1 | Top-1 accuracy: 42.1 -> 43.5 | Improved |
| add end-to-end joint fine-tuning of the downscaler (ablation) | TSM with SelfC · Something V1 | Top-1 accuracy: 43.5 -> 44.6 | Improved |
Closest papers by meaning
- Video Rescaling Networks With Joint Optimization Strategies for Downscaling and Upscaling (CVPR 2021 · similarity 0.87)
- Self-Supervised Video Representation Learning by Context and Motion Decoupling (CVPR 2021 · similarity 0.85)
- Deep Hierarchical Video Compression (AAAI 2024 · similarity 0.83)
- Visual Representation Learning with Stochastic Frame Prediction (ICML 2024 · similarity 0.83)
- TTVidT: Decoupling the Temporal Axis for Efficient Motion-Centric Video Pretraining (NeurIPS 2026 · similarity 0.83)
- Boosting Neural Representations for Videos with a Conditional Decoder (CVPR 2024 · similarity 0.83)
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.
