Research map / Video / Video segmentation and tracking
Learning to Generate Long-term Future via Hierarchical Prediction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 hierarchical pose prediction and pose-conditioned frame generation instead of recursive pixel prediction | Long-term video prediction · Penn Action | worker preference over ConvLSTM: 82.4% | Helped |
| Use hierarchical pose prediction and pose-conditioned frame generation instead of optical-flow pixel warping | Long-term video prediction · Penn Action | worker preference over Optical Flow: 86.1% | Helped |
| Use hierarchical pose prediction and pose-conditioned frame generation instead of recursive pixel prediction | Long-term video prediction · Human3.6M | worker preference over ConvLSTM: 70.3% | Helped |
| Use hierarchical pose prediction and pose-conditioned frame generation instead of optical-flow pixel warping | Long-term video prediction · Human3.6M | worker preference over Optical Flow: 72.3% | Helped |
| Use hierarchical pose prediction and pose-conditioned frame generation instead of Convolutional LSTM | Penn Action generated videos · Penn Action | Combined action recognition accuracy: 46.4% -> 59.0% | Helped |
| Add copying of background pixels from the last observed frame using predicted pose masks (ablation) | Ours · Human3.6M | PSNR | Mixed |
| Provide the image generator with ground-truth future poses instead of predicted future poses (ablation) | Ours · Penn Action | PSNR | Helped |
| Provide the image generator with ground-truth future poses instead of predicted future poses (ablation) | Ours · Human3.6M | PSNR | Helped |
Closest papers by meaning
- Revisiting Hierarchical Approach for Persistent Long-Term Video Prediction (ICLR 2021 · similarity 0.93)
- Hierarchical Long-term Video Prediction without Supervision (ICML 2018 · similarity 0.91)
- Decomposing Motion and Content for Natural Video Sequence Prediction (ICLR 2017 · similarity 0.89)
- Predicting Future Frames Using Retrospective Cycle GAN (CVPR 2019 · similarity 0.88)
- Visual Representation Learning with Stochastic Frame Prediction (ICML 2024 · similarity 0.86)
- FutureHuman3D: Forecasting Complex Long-Term 3D Human Behavior from Video Observations (CVPR 2024 · 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.