Research map / Detection and segmentation / Tracking
Type-to-Track: Retrieve Any Object via Prompt-based Tracking
Pha Nguyen, Kha Gia Quach, Kris M. Kitani, Khoa Luu
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 two-stage MDETR+TrackFormer with single-stage MENDER | MENDER · GroOT TAO, retr prompts | CA-MOTA: 25.70 -> 32.90 | Improved |
| replace two-stage MDETR+TrackFormer with MENDER | MENDER · GroOT MOT17, cap prompts | CA-MOTA: 44.80 -> 59.50 | Improved |
| replace two-stage baseline with MENDER | MENDER · GroOT MOT17, def prompts | FPS: 2.2 -> 10.3 | Improved |
| replace two-stage baseline with MENDER | MENDER · GroOT TAO, def prompts | CA-MOTA: 14.60 -> 16.80 | Improved |
| use simplified correlation instead of full third-order tensor (ablation) | MENDER · GroOT MOT17, cap prompts | FPS: 3.4 -> 7.8 | Improved |
| use simplified correlation instead of full third-order tensor (ablation) | MENDER · GroOT TAO, retr prompts | FPS: 7.6 -> 11.5 | Improved |
| use simplified correlation instead of full tensor (ablation) | MENDER · GroOT MOT17, cap prompts | CA-MOTA: 58.20 -> 59.50 | Improved |
| use simplified correlation instead of full tensor (ablation) | MENDER · GroOT TAO, retr prompts | CA-MOTA: 32.40 -> 32.90 | Improved |
| use simplified correlation instead of full tensor (ablation) | MENDER · GroOT MOT20, def prompts | CA-MOTA: 72.90 -> 72.10 | Hurt |
| use simplified correlation instead of full tensor (ablation) | MENDER · GroOT MOT17, cap prompts | MT: 289 -> 201 | Hurt |
| use detailed captions (cap) instead of category definitions (def) as prompt | MENDER · GroOT MOT17 | CA-MOTA: 67.30 -> 59.50 | Hurt |
| use detailed captions (cap) instead of category definitions (def) as prompt | MENDER · GroOT TAO | CA-MOTA: 16.80 -> 20.70 | Improved |
Closest papers by meaning
- Generalizing Multiple Object Tracking to Unseen Domains by Introducing Natural Language Representation (AAAI 2023 · similarity 0.91)
- Towards More Flexible and Accurate Object Tracking With Natural Language: Algorithms and Benchmark (CVPR 2021 · similarity 0.90)
- Beyond Explicit Language: Plug-and-Play Visual-to-Linguistic Modeling Toward General Object Tracking (CVPR 2026 · similarity 0.89)
- Tracking by Natural Language Specification (CVPR 2017 · similarity 0.89)
- OVTR: End-to-End Open-Vocabulary Multiple Object Tracking with Transformer (ICLR 2025 · similarity 0.88)
- ChatTracker: Enhancing Visual Tracking Performance via Chatting with Multimodal Large Language Model (NeurIPS 2024 · similarity 0.87)
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.
