Research map / 3D vision / Human pose estimation
Hierarchical Video Prediction Using Relational Layouts for Human-Object Interactions
Navaneeth Bodla, Gaurav Shrivastava, Rama Chellappa, Abhinav Shrivastava
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 layouts containing both pose and object locations instead of pose-only structure | HORN · Bimanual | FID: 69.92 -> 35.82 | Improved |
| Add relational reasoning between human pose and objects (ablation) | HORN · Bimanual | FID: 65.73 -> 35.82 | Improved |
| Add relational reasoning between human pose and objects (ablation) | HORN · Bimanual | mAP: 55.46 -> 59.04 | Improved |
| Add prior layouts to video generation (ablation) | HORN · Bimanual | FID: 42.97 -> 35.82 | Improved |
| Condition layout generation on the first frame rather than the previous generated frame (ablation) | HORN · Bimanual | FID: 140.80 -> 35.82 | Improved |
| Add frame GAN to pixel and feature losses (ablation) | HORN · Bimanual | FID: 49.05 -> 40.44 | Improved |
| Add frame GAN to pixel and feature losses (ablation) | HORN · Bimanual | mAP: 53.54 -> 57.21 | Improved |
| Add object GAN to pixel, feature, and frame GAN losses (ablation) | HORN · Bimanual | mAP: 57.21 -> 58.67 | Improved |
| Add video GAN to pixel, feature, frame, and object GAN losses (ablation) | HORN · Bimanual | mAP: 58.67 -> 59.04 | Improved |
| Add video GAN to pixel, feature, frame, and object GAN losses (ablation) | HORN · Bimanual | FID: 38.30 -> 35.83 | Improved |
Closest papers by meaning
- Attend and Interact: Higher-Order Object Interactions for Video Understanding (CVPR 2018 · similarity 0.88)
- Unsupervised Learning for Physical Interaction through Video Prediction (NeurIPS 2016 · similarity 0.88)
- Learning Asynchronous and Sparse Human-Object Interaction in Videos (CVPR 2021 · similarity 0.88)
- The Pose Knows: Video Forecasting by Generating Pose Futures (ICCV 2017 · similarity 0.88)
- Learning 3D Human Dynamics From Video (CVPR 2019 · similarity 0.87)
- FutureHuman3D: Forecasting Complex Long-Term 3D Human Behavior from Video Observations (CVPR 2024 · similarity 0.87)
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.
