Research map / Video / Video segmentation and tracking
Probabilistic Video Prediction From Noisy Data With a Posterior Confidence
Yunbo Wang, Jiajun Wu, Mingsheng Long, Joshua B. Tenenbaum
Check 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 |
|---|---|---|---|
| Use BP-Net instead of SVG-LP | Probabilistic video prediction from noisy inputs · Moving MNIST, perceptual uncertainty | SSIM: 0.789 -> 0.810 | Improved |
| Use BP-Net instead of SVG-LP | Probabilistic video prediction from noisy inputs · Moving MNIST, entangled perceptual and dynamics uncertainty | SSIM: 0.757 -> 0.788 | Improved |
| Use BP-Net instead of SVG-LP | Probabilistic video prediction from noisy inputs · KTH, noisy inputs | SSIM: 0.775 -> 0.792 | Improved |
| Add convolutional recurrence in the prediction module (ablation) | BP-Net on Moving MNIST · Moving MNIST, entangled perceptual and dynamics uncertainty | SSIM: 0.765 -> 0.788 | Improved |
| Add convolutional recurrence in the prediction module (ablation) | BP-Net on Moving MNIST · Moving MNIST, entangled perceptual and dynamics uncertainty | MSE: 63.4 -> 58.5 | Improved |
| Add random noise vectors to the measurement module (ablation) | BP-Net on Moving MNIST · Moving MNIST, entangled perceptual and dynamics uncertainty | SSIM: 0.782 -> 0.788 | Improved |
| Add random noise vectors to the measurement module (ablation) | BP-Net on Moving MNIST · Moving MNIST, entangled perceptual and dynamics uncertainty | Highest particle weight: 0.99 -> 0.23 | Improved |
| Train denoising and prediction as separate stages rather than end-to-end (ablation) | Video prediction from noisy frames · Moving MNIST, entangled perceptual and dynamics uncertainty | SSIM: 0.756 vs 0.788 | Hurt |
| Train denoising and prediction as separate stages rather than end-to-end (ablation) | Video prediction from noisy frames · Moving MNIST, entangled perceptual and dynamics uncertainty | MSE: 68.0 vs 58.5 | Hurt |
Closest papers by meaning
- A Hierarchical Variational Neural Uncertainty Model for Stochastic Video Prediction (ICCV 2021 · similarity 0.87)
- Stochastic Variational Video Prediction (ICLR 2018 · similarity 0.87)
- VideoFlow: A Conditional Flow-Based Model for Stochastic Video Generation (ICLR 2020 · similarity 0.86)
- Flexible Spatio-Temporal Networks for Video Prediction (CVPR 2017 · similarity 0.86)
- STDiff: Spatio-Temporal Diffusion for Continuous Stochastic Video Prediction (AAAI 2024 · similarity 0.84)
- DFDNet: Disentangling and Filtering Dynamics for Enhanced Video Prediction (AAAI 2025 · similarity 0.84)
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.
