Research map / Video / Video segmentation and tracking
3R-INN: How to be climate friendly while consuming/delivering videos?
ZOUBIDA AMEUR, Claire-Helene Demarty, Olivier LE MEUR, Daniel Menard
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 conditional latent encoding block (condition latent mean/variance on LR image) (ablation) | 3R-INN · DIV2K val grainy HR | JSD-NSS: 0.0096 -> 0.0090 | Improved |
| add conditional latent encoding block (ablation) | 3R-INN · DIV2K val clean LR | PSNR: 38.62 -> 38.71 | No clear effect |
| increase lambda1 weight of forward fidelity loss to 40 (ablation) | 3R-INN · DIV2K val clean LR | PSNR: 38.71 -> 39.30 | Improved |
| increase lambda1 to 40 (ablation) | 3R-INN · DIV2K val clean HR | PSNR: 35.53 -> 35.41 | Hurt |
| increase film grain latent dimension dim(z_G) beyond 1 (ablation) | 3R-INN · DIV2K val | clean LR/HR PSNR: 39.45 -> 39.30 LR PSNR for dim 1 vs 2 | Hurt |
| optimize RGBW OLED power model instead of luminance model P_Y | 3R-INN energy-aware task · DIV2K, R=20% | PSNR-Y: 27.32 -> 38.02 | Improved |
| use single invertible network vs IRN for rescaling plus grain removal | 3R-INN · DIV2K clean LR | PSNR: 39.06 -> 39.63 | Improved |
| 3R-INN vs IRN on clean HR reconstruction | 3R-INN · DIV2K clean HR | PSNR: 36.53 -> 35.43 | Hurt |
| encode low-res grain-free energy-aware video instead of HR grainy | video chain (RaceHorses, QP22, R=20%) · RaceHorses | energy savings: 78% encoding, 3% delivery, ca. 77% decoding | Improved |
| energy-aware reduction rate R on OLED display | 3R-INN · RaceHorses on LG-42C2 OLED | display power reduction: 6.8%, 21.5%, 33.3%, 44.2% for R=5/20/40/60% | Improved |
Closest papers by meaning
- LinGen: Towards High-Resolution Minute-Length Text-to-Video Generation with Linear Computational Complexity (CVPR 2025 · similarity 0.80)
- TeTriRF: Temporal Tri-Plane Radiance Fields for Efficient Free-Viewpoint Video (CVPR 2024 · similarity 0.80)
- CE-NAS: An End-to-End Carbon-Efficient Neural Architecture Search Framework (NeurIPS 2024 · similarity 0.79)
- Compressing Streamable Free-Viewpoint Videos to 0.1 MB per Frame (AAAI 2025 · similarity 0.79)
- E4: Energy-Efficient DNN Inference for Edge Video Analytics via Early Exiting and DVFS (AAAI 2025 · similarity 0.79)
- AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration (ICML 2025 · similarity 0.79)
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.
