Research map / Video / Video segmentation and tracking
Efficient Neural Video Representation with Temporally Coherent Modulation
Seungjun Shin, Suji Kim, Dokwan Oh
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 temporally coherent modulation with alignment flow (NVTM instead of NVP) | grid-based video INR · UVG (Dynamic) | PSNR: 39.00 -> 40.54 | Improved |
| add temporally coherent modulation with alignment flow (NVTM instead of NVP) | grid-based video INR · MCL-JCV (Dynamic) | PSNR: 39.55 -> 41.39 | Improved |
| add adaptive normalization (ablation) | NVTM · UVG (Dynamic) | PSNR: 40.24 -> 40.36 | Improved |
| add static feature grid (ablation) | NVTM · UVG (Dynamic) | PSNR: 40.36 -> 40.54 | Improved |
| use latent as modulation rather than direct input (ablation) | NVTM · UVG (Dynamic) | PSNR: 38.90 -> 40.54 | Improved |
| use two neighboring GOP latent grids P={0,1} instead of one (ablation) | NVTM · UVG (Dynamic) | PSNR: 39.45 -> 40.54 | Improved |
| reduce GOP size from 10 to 5 (ablation) | NVTM · UVG per-sequence | PSNR: YachtRide 38.88 -> 39.57; Bosphorus 43.30 -> 43.27 | Mixed |
| increase GOP size from 10 to 60 (ablation) | NVTM · UVG YachtRide | PSNR: 38.88 -> 32.53 | Hurt |
| shrink temporal grid size T below video length with fixed total parameters (ablation) | NVP · UVG 600 frames | PSNR: 36.34 -> 29.47 | Hurt |
| use NVTM instead of NVP on long videos (2400 frames) | grid-based video INR · UVG 2400 frames | PSNR: 40.21 -> 41.05 | Improved |
| use NVTM instead of NVP for frame interpolation | grid-based video INR · UVG (Dynamic) frame interpolation | PSNR: 23.88 -> 30.49 | Improved |
| use NVTM instead of HNeRV at long encoding time (~60min) on UVG | video INR at 0.1bpp · UVG (Dynamic), ~60min | PSNR: 32.51 vs 31.85 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- HNeRV — HNeRV: A Hybrid Neural Representation for Videos (CVPR 2023) · better on 3 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- NVRC: Neural Video Representation Compression (NeurIPS 2024 · similarity 0.91)
- Combining Frame and GOP Embeddings for Neural Video Representation (CVPR 2024 · similarity 0.91)
- Boosting Neural Representations for Videos with a Conditional Decoder (CVPR 2024 · similarity 0.90)
- DS-NeRV: Implicit Neural Video Representation with Decomposed Static and Dynamic Codes (CVPR 2024 · similarity 0.90)
- Progressive Fourier Neural Representation for Sequential Video Compilation (ICLR 2024 · similarity 0.88)
- Towards Practical Real-Time Neural Video Compression (CVPR 2025 · 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.
