Research map / Video / Action recognition
Temporal Hallucinating for Action Recognition With Few Still Images
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 |
|---|---|---|---|
| Add hybrid video memory with temporal hallucinating to the standard image-only GP (ablation) | GP action recognition · WEB101, 1-image | classification accuracy: 26.6 -> 35.4 | Helped |
| Add hybrid video memory with temporal hallucinating to the standard image-only GP (ablation) | GP action recognition · VOC, 1-image | classification accuracy: 39.0 -> 42.2 | Helped |
| Add hybrid video memory with temporal hallucinating to the standard image-only GP (ablation) | GP action recognition · DIFF20, 1-image | classification accuracy: 57.4 -> 60.2 | Helped |
| Add spatial-temporal prediction fusion to temporal predicting (ablation) | HVM · WEB101, 1-image | classification accuracy: 33.2 -> 35.4 | Helped |
| Add spatial-temporal prediction fusion to temporal predicting (ablation) | HVM · VOC, 1-image | classification accuracy: 42.0 -> 42.2 | Helped |
| Add spatial-temporal prediction fusion to temporal predicting (ablation) | HVM · DIFF20, 1-image | classification accuracy: 59.1 -> 60.2 | Helped |
| Select relevant videos with the video-selection module instead of selecting them randomly (ablation) | HVM temporal predicting · VOC, 1-image | classification accuracy: 38.2 -> 42.0 | Helped |
| Select relevant videos with the video-selection module instead of selecting them randomly (ablation) | HVM temporal predicting · DIFF20, 1-image | classification accuracy: 53.9 -> 59.1 | Helped |
| Use temporal predicting with randomly selected video memory (ablation) | HVM · VOC, 1-image | classification accuracy: 39.0 -> 38.2 | Hurt |
| Use temporal predicting with randomly selected video memory (ablation) | HVM · DIFF20, 1-image | classification accuracy: 57.4 -> 53.9 | Hurt |
| Replace the linear kernel with the nonlinear kernel (ablation) | HVM · WEB101, 1-image | classification accuracy: 35.4 -> 37.5 | Helped |
| Replace the linear kernel with the nonlinear kernel (ablation) | HVM · VOC, 1-image | classification accuracy: 42.2 -> 41.2 | Hurt |
Closest papers by meaning
- Memory-Augmented Temporal Dynamic Learning for Action Recognition (AAAI 2019 · similarity 0.89)
- MASH-VLM: Mitigating Action-Scene Hallucination in Video-LLMs through Disentangled Spatial-Temporal Representations (CVPR 2025 · similarity 0.88)
- MHBench: Demystifying Motion Hallucination in VideoLLMs (AAAI 2025 · similarity 0.87)
- VidHalluc: Evaluating Temporal Hallucinations in Multimodal Large Language Models for Video Understanding (CVPR 2025 · similarity 0.87)
- Mitigating Hallucination in VideoLLMs via Temporal-Aware Activation Engineering (NeurIPS 2025 · similarity 0.86)
- ActFusion: a Unified Diffusion Model for Action Segmentation and Anticipation (NeurIPS 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.