Research map / Detection and segmentation / CNN features
Online Video Object Detection Using Association LSTM
Yongyi Lu, Cewu Lu, Chi-Keung Tang
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 |
|---|---|---|---|
| add LSTM regression of boxes and scores on top of SSD (ablation) | Baseline 1 (SSD) · Youtube-Objects | mAP: 66.21 -> 69.08 | Improved |
| add reversed-sequence augmentation (ablation) | Association LSTM · Youtube-Objects | mAP: 69.08 -> 70.2 | Improved |
| add the association error term (ablation) | Association LSTM · Youtube-Objects | mAP: 70.2 -> 72.14 | Improved |
| combine regression and association vs SSD alone | Baseline 1 · Youtube-Objects | mAP: 66.21 -> 72.14 | Improved |
| use association LSTM instead of GRU context refinement | Context [27] · Youtube-Objects | mAP: 68.73 -> 72.14 | Improved |
| use association LSTM instead of keypoint tracker post-processing | Baseline 2 · Youtube-Objects | mAP: 70.43 -> 72.14 | Improved |
| use association LSTM instead of CDT on tracking | CDT · MOTChallenge validation | MOTA: 39.9 -> 38.6 | Hurt |
| use association LSTM instead of CDT on mostly-lost | CDT · MOTChallenge validation | ML: 47.4 -> 46.8 | Improved |
| use association LSTM instead of CDT on ID switches | CDT · MOTChallenge validation | IDS: 95 -> 154 | Hurt |
| use association LSTM instead of CDT on false positives | CDT · MOTChallenge validation | FP: 914 -> 788 | Improved |
Closest papers by meaning
- Mobile Video Object Detection With Temporally-Aware Feature Maps (CVPR 2018 · similarity 0.87)
- Temporal Dynamic Graph LSTM for Action-Driven Video Object Detection (ICCV 2017 · similarity 0.87)
- Leveraging Long-Range Temporal Relationships Between Proposals for Video Object Detection (ICCV 2019 · similarity 0.87)
- Bridging Images and Videos: A Simple Learning Framework for Large Vocabulary Video Object Detection (ECCV 2022 · similarity 0.86)
- LSTFE-Net:Long Short-Term Feature Enhancement Network for Video Small Object Detection (CVPR 2023 · similarity 0.86)
- "A Simple Approach and Benchmark for 21,000-Category Object Detection" (ECCV 2022 · 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.
