Research map / Detection and segmentation / Tracking
Referring Multi-Object Tracking with Diverse Dynamic Expressions
Yani Zhang, Dongming Wu, Wencheng Han, Xingping Dong, Shengcai Liao, Bo Du
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 Temporal Enhancement module to TransRMOT baseline (ablation) | TransRMOT baseline · Refer-KITTI-V2 | HOTA: 31.00 -> 34.46 | Improved |
| add Track Refinement module (ablation) | TempRMOT · Refer-KITTI-V2 | HOTA: 34.46 -> 35.04 | Improved |
| add temporal decoder only (ablation) | temporal enhancement module · Refer-KITTI-V2 | HOTA: 31.00 -> 33.52 | Improved |
| add object decoder only (ablation) | temporal enhancement module · Refer-KITTI-V2 | HOTA: 31.00 -> 32.90 | Improved |
| add positional embeddings to temporal+object decoders (ablation) | temporal enhancement module · Refer-KITTI-V2 | HOTA: 34.43 -> 35.04 | Improved |
| increase training memory length from 3 to 5 (ablation) | TempRMOT · Refer-KITTI-V2 | HOTA: 33.64 -> 34.72 | Improved |
| increase inference memory length from 5 to 8 (trained with 5) (ablation) | TempRMOT · Refer-KITTI-V2 | HOTA: 34.72 -> 35.04 | Improved |
| use proposed temporal module instead of GRU memory (ablation) | TempRMOT · Refer-KITTI-V2 | HOTA: 31.88 -> 35.04 | Improved |
| use proposed temporal module instead of convolution (ablation) | TempRMOT · Refer-KITTI-V2 | AssA: 44.65 -> 53.58 | Improved |
| add LLM-generated expressions to human expressions (ablation) | TempRMOT · Refer-KITTI-V2 | HOTA: 27.69 -> 35.04 | Improved |
| add LLM-generated expressions on independently written test expressions (ablation) | TempRMOT · Independently written expressions | HOTA: 30.09 -> 37.90 | Improved |
| train on Refer-KITTI-V2 instead of Refer-KITTI for cross-dataset test (ablation) | TempRMOT · GroOT (MOT17 subset) | HOTA: 6.75 -> 12.01 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- iKUN — iKUN: Speak to Trackers without Retraining (CVPR 2024) · better on 2 of 2 shared results
- MOTRv2 — MOTRv2: Bootstrapping End-to-End Multi-Object Tracking by Pretrained Object Detectors (CVPR 2023) · better on 1 of 1 shared result
- ByteTrack — ByteTrack: Multi-Object Tracking by Associating Every Detection Box (ECCV 2022) · better on 1 of 1 shared result
- FairMOT — FairMOT: On the Fairness of Detection and Re-identification in Multiple Object Tracking (IJCV 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Referring Multi-Object Tracking (CVPR 2023 · similarity 0.95)
- Language Decoupling with Fine-grained Knowledge Guidance for Referring Multi-object Tracking (ICCV 2025 · similarity 0.92)
- YesTrack: Referring Multi-Object Tracking via MLLM-based Yes/No Reasoning (ECCV 2026 · similarity 0.92)
- Cross-View Referring Multi-Object Tracking (AAAI 2025 · similarity 0.90)
- iKUN: Speak to Trackers without Retraining (CVPR 2024 · similarity 0.88)
- HFF-Tracker: A Hierarchical Fine-grained Fusion Tracker for Referring Multi-Object Tracking (AAAI 2025 · 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.
