Research map / Video / Action recognition
Im2Flow: Motion Hallucination From Static Images for Action Recognition
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 |
|---|---|---|---|
| predict flow with Im2Flow rather than Walker et al.'s method | dense optical flow prediction · UCF-101 | EPE: 2.391 -> 2.210 | Helped |
| predict flow with Im2Flow rather than Walker et al.'s method | dense optical flow prediction · UCF-101 | Direction Similarity: 0.003 -> 0.143 | Helped |
| predict flow with Im2Flow rather than Walker et al.'s method | dense optical flow prediction · UCF-101 | Orientation Similarity: 0.661 -> 0.699 | Helped |
| add inferred motion from Ours-HMDB to the appearance stream | static-image action recognition · UCF-Static | accuracy: 63.6 -> 65.5 | Helped |
| add inferred motion from Ours-UCF to the appearance stream | static-image action recognition · HMDB-Static | accuracy: 35.1 -> 37.1 | Helped |
| add inferred motion from Ours-UCF+HMDB to the appearance stream | static-image action recognition · PennAction | accuracy: 73.1 -> 74.5 | Helped |
| add inferred motion from Ours-UCF+HMDB to the appearance stream | static-image action recognition · Willow | accuracy: 65.1 -> 67.5 | Helped |
| add inferred motion from Ours-UCF+HMDB to the appearance stream | static-image action recognition · Stanford10 | accuracy: 81.3 -> 82.3 | Helped |
| add inferred motion from Ours-UCF+HMDB to the appearance stream | static-image action recognition · PASCAL2012 Actions | accuracy: 65.0 -> 66.1 | Helped |
| add inferred motion to the appearance stream | static-image dynamic scene recognition · static-YUP++ | accuracy: 74.3 -> 78.2 | Helped |
| add inferred motion to the appearance stream | static-image dynamic scene recognition · static-YUP++ | mAP: 79.3 -> 82.3 | Helped |
| use ResNet-50 rather than AlexNet as the base network | Im2Flow recognition model · Willow | mAP: 74.0 -> 90.5 | Helped |
Closest papers by meaning
- Dance With Flow: Two-In-One Stream Action Detection (CVPR 2019 · similarity 0.88)
- MARS: Motion-Augmented RGB Stream for Action Recognition (CVPR 2019 · similarity 0.88)
- Temporal Hallucinating for Action Recognition With Few Still Images (CVPR 2018 · similarity 0.87)
- H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning (NeurIPS 2026 · similarity 0.87)
- Flow Snapshot Neurons in Action: Deep Neural Networks Generalize to Biological Motion Perception (NeurIPS 2024 · similarity 0.85)
- MHBench: Demystifying Motion Hallucination in VideoLLMs (AAAI 2025 · 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.