Research map / Video / Video segmentation and tracking
Generating the Future With Adversarial Transformers
Carl Vondrick, Antonio Torralba
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 adversarial transformation prediction instead of regression transformation prediction | video future generation · Flickr video future generation | human preference rate: 55.6 | Improved |
| use transformations rather than direct pixel-intensity prediction with adversarial learning (ablation) | video future generation · Flickr video future generation | human preference rate: 61.2 | Improved |
| use transformations rather than direct pixel-intensity prediction with regression (ablation) | video future generation · Flickr video future generation | human preference rate: 55.1 | Improved |
| add transformation prediction to the adversarial model (ablation) | object recognition initialization · PASCAL VOC 2007 | mAP: 29.7 -> 32.0 | Improved |
| add transformation prediction to the adversarial model (ablation) | object recognition initialization · PASCAL VOC 2012 | mAP: 33.3 -> 38.1 | Improved |
| add transformation prediction to the regression model (ablation) | object recognition initialization · PASCAL VOC 2007 | mAP: 30.0 -> 32.6 | Improved |
| add transformation prediction to the regression model (ablation) | object recognition initialization · PASCAL VOC 2012 | mAP: 33.6 -> 38.8 | Improved |
| replace adversarial learning with regression while retaining transformations (ablation) | object recognition initialization · PASCAL VOC 2007 | mAP: 32.6 vs 32.0 | No clear effect |
| replace adversarial learning with regression while retaining transformations (ablation) | object recognition initialization · PASCAL VOC 2012 | mAP: 38.8 vs 38.1 | No clear effect |
Closest papers by meaning
- Learning Temporal Transformations from Time-Lapse Videos (ECCV 2016 · similarity 0.90)
- Predicting the Future: A Jointly Learnt Model for Action Anticipation (ICCV 2019 · similarity 0.87)
- Learning to Generate Long-term Future via Hierarchical Prediction (ICML 2017 · similarity 0.87)
- Learning the Predictability of the Future (CVPR 2021 · similarity 0.87)
- Advancing Semantic Future Prediction through Multimodal Visual Sequence Transformers (CVPR 2025 · similarity 0.86)
- Towards Understanding Future: Consistency Guided Probabilistic Modeling for Action Anticipation (AAAI 2024 · similarity 0.85)
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.
