Research map / Video / Video segmentation and tracking
MLVC: A Multi-platform Learned Video Codec for Real-World Deployment
Tanel Pärnamaa, Martin Lumiste, Ardi Loot, Evgenii Indenbom, Andrei Znobishchev, Ando Saabas
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 scale sending through the hyperprior (vs no adaptive priors) (ablation) | MLVC · VCD-TH XP | BD-rate (PSNR): 718.1 -> 101.8 | Improved |
| add scale sending with WSiLU relative to DCVC-RT cross-platform failure | MLVC · Avg SP | BD-rate (PSNR): -69.6 -> -54.5 | Hurt |
| replace WSiLU with LeakyReLU (ablation) | MLVC · VCD-TH XP | BD-rate (PSNR): 101.8 -> 6.7 | Improved |
| replace WSiLU with LeakyReLU (ablation) | MLVC · Avg SP | BD-rate (PSNR): -54.5 -> -49.2 | Hurt |
| add memory module (gated) (ablation) | MLVC · Avg SP | BD-rate (PSNR): -49.2 -> -52.1 | Improved |
| add memory module (gated) (ablation) | MLVC · VCD-TH XP | BD-rate (PSNR): 6.7 -> -21.9 | Improved |
| replace LeakyReLU with ReGLU (ablation) | MLVC · Avg SP | BD-rate (PSNR): -52.1 -> -56.6 | Improved |
| replace LeakyReLU with ReGLU (ablation) | MLVC · VCD-TH XP | BD-rate (PSNR): -21.9 -> -39.8 | Improved |
| add I-frame dropout (ablation) | MLVC · Avg SP | BD-rate (PSNR): -41.1 -> -46.1 | Improved |
| add long-term reference recovery with IP 1024 (ablation) | MLVC · VCD-TH XP | BD-rate (PSNR): -44.9 -> -58.7 | Improved |
| add long-term reference recovery with IP 1024 (ablation) | MLVC · Avg SP | BD-rate (PSNR): -46.1 -> -52.0 | Improved |
| add LTR period 32 at IP 128 (vs none) | MLVC · GPU-NPU cross-device | BD-rate: -47.0 -> -41.4 | Improved |
Closest papers by meaning
- Towards Practical Real-Time Neural Video Compression (CVPR 2025 · similarity 0.90)
- Neural Video Compression with Feature Modulation (CVPR 2024 · similarity 0.89)
- PNVC: Towards Practical INR-based Video Compression (AAAI 2025 · similarity 0.88)
- DeNC: Unleash Neural Codecs in Video Streaming with Diffusion Enhancement (AAAI 2025 · similarity 0.87)
- Neural Video Compression with Context Modulation (CVPR 2025 · similarity 0.86)
- When MLLMs Meets Compression Distortion: A Coding Paradigm Tailored to MLLMs (ICLR 2026 · similarity 0.86)
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.
