Research map / Video / Video segmentation and tracking
Video Compression With Rate-Distortion Autoencoders
Amirhossein Habibian, Ties van Rozendaal, Jakub M. Tomczak, Taco S. Cohen
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 |
|---|---|---|---|
| condition the autoregressive prior on the previous latent frame (ablation) | frame-based prior · SDV (trained on Kinetics) | rate-distortion | Improved |
| add a Conv-GRU to the prior instead of last-frame conditioning (ablation) | video prior (last frame) · SDV (trained on Kinetics) | rate-distortion | No clear effect |
| use a 3D spatio-temporal autoencoder instead of a 2D frame autoencoder (ablation) | frame AE · SDV (trained on Kinetics) | rate-distortion | Improved |
| weight rate and distortion losses toward person regions (alpha 0.95) | non-semantic model · Human Activity | MS-SSIM foreground vs background | Mixed |
| finetune generic model on the target driving domain | generic Kinetics model · Dynamics | rate-distortion | Improved |
| train on channel-wise concatenation of four camera streams instead of single streams | unimodal model · MHAD | bits per pixel: 0.035 -> 0.007 bpp vs HEVC | Improved |
Closest papers by meaning
- Insights From Generative Modeling for Neural Video Compression (TPAMI 2023 · similarity 0.92)
- Deep Generative Video Compression (NeurIPS 2019 · similarity 0.92)
- Joint Autoregressive and Hierarchical Priors for Learned Image Compression (NeurIPS 2018 · similarity 0.91)
- Lossy Image Compression with Compressive Autoencoders (ICLR 2017 · similarity 0.91)
- Overfitting for Fun and Profit: Instance-Adaptive Data Compression (ICLR 2021 · similarity 0.90)
- Variational image compression with a scale hyperprior (ICLR 2018 · similarity 0.90)
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.
