Research map / Video / Video segmentation and tracking
Video Imprint
Zhanning Gao, Le Wang, Nebojša Jojić, Zhenxing Niu, Nanning Zheng, Gang Hua
Read the paperCheck 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 |
|---|---|---|---|
| replace sum-aggregation with counting grid aggregation (CGA) | video representation (alex+res) · EVVE retrieval | mAP: 47.3 -> 52.3 | Improved |
| replace sum-aggregation with epitome aggregation (EPA) | video representation (alex+res) · EVVE retrieval | mAP: 47.3 -> 52.1 | Improved |
| replace sum-aggregation with CGA | AlexNet features · EVVE retrieval | mAP: 38.3 -> 42.6 | Improved |
| replace sum-aggregation with CGA | ResNet-50 features · EVVE retrieval | mAP: 46.6 -> 51.2 | Improved |
| concatenate AlexNet and ResNet-50 descriptors | CGA · EVVE retrieval | mAP: 51.2 -> 52.3 | Improved |
| replace sum-aggregation with CGA | video representation · EVVE+100K retrieval | mAP: 38.7 -> 42.9 | Improved |
| add difference-of-neighborhood query expansion | CGA-(alex+res) · EVVE retrieval | mAP: 52.3 -> 60.1 | Improved |
| use efficient two-step scheme (d=64) for epitome learning (ablation) | epitome · EVVE | running time | Improved |
| increase counting grid size beyond 24 (ablation) | TCG/epitome · EVVE retrieval | mAP | No clear effect |
| add reasoning network re-weighting to CGA | CGA-(vgg+res) · CCV | mAP: 79.1 -> 79.9 | Improved |
| add reasoning network re-weighting to CGA | CGA-(vgg+res) · MED14 | mAP: 33.7 -> 36.9 | Improved |
| fuse motion and audio features | CG-RNet-(vgg+res) · CCV | mAP: 79.9 -> 87.1 | Improved |
Closest papers by meaning
- ER3: A Unified Framework for Event Retrieval, Recognition and Recounting (CVPR 2017 · similarity 0.96)
- Video Imprint Segmentation for Temporal Action Detection in Untrimmed Videos (AAAI 2019 · similarity 0.83)
- Video Event Extraction with Multi-View Interaction Knowledge Distillation (AAAI 2024 · similarity 0.82)
- Precise Event Spotting in Sports Videos: Solving Long-Range Dependency and Class Imbalance (CVPR 2025 · similarity 0.81)
- OmniViD: A Generative Framework for Universal Video Understanding (CVPR 2024 · similarity 0.81)
- A Unified Reasoning Framework for Holistic Zero-Shot Video Anomaly Analysis (NeurIPS 2025 · similarity 0.81)
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.
