Research map / Vision-language and multimodal / Human and embodied modeling
SeqTR: A Simple Yet Universal Network for Visual Grounding
Chaoyang Zhu, Yiyi Zhou, Yunhang Shen, Gen Luo, Xingjia Pan, Mingbao Lin, Chao Chen, Liujuan Cao, Xiaoshuai Sun, Rongrong Ji
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 |
|---|---|---|---|
| use max pooling of word features instead of mean pooling (ablation) | SeqTR · RefCOCO val | Prec@0.5 (%): 79.73 -> 80.07 | Improved |
| use max pooling instead of final bi-GRU hidden state (ablation) | SeqTR · RefCOCO+ val | Prec@0.5 (%): 67.46 -> 68.31 | Improved |
| raise 1st token weight to 1.5 (ablation) | SeqTR · RefCOCO+ val | Prec@0.5 (%): 68.31 -> 68.63 | Improved |
| raise first token weight to 3 (ablation) | SeqTR · RefCOCO+ val | Prec@0.5 (%): 68.31 -> 67.81 | Hurt |
| raise 2nd and 3rd token weights to 2 (ablation) | SeqTR · RefCOCOg val-u | Prec@0.5 (%): 69.95 -> 69.51 | Hurt |
| add multi-task REC+RES supervision (ablation) | SeqTR · RefCOCO val | RES mIoU (%): 64.20 -> 62.93 | Hurt |
| add multi-task REC+RES supervision (ablation) | SeqTR · RefCOCO+ val | RES mIoU (%): 52.22 -> 53.65 | Improved |
| add multi-task REC+RES supervision (ablation) | SeqTR · RefCOCOg val-u | RES mIoU (%): 53.25 -> 53.25 | No clear effect |
| use uniform instead of center-based mask contour sampling (ablation) | SeqTR · RES upper bound | mIoU upper bound: 91.58 -> 95.57 | Improved |
| add large-scale pre-training | SeqTR · RefCOCO val (RES) | mIoU (%): 67.26 -> 71.70 | Improved |
| add large-scale pre-training | SeqTR · RefCOCO+ val (RES) | mIoU (%): 54.14 -> 63.04 | Improved |
| sample with nucleus top-p 0.6 instead of argmax (ablation) | SeqTR · RefCOCO val (RES) | mIoU (%): 67.26 -> 66.15 | Hurt |
Later papers that compared with this method
1 later accepted paper reports SeqTR as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Parallel Vertex Diffusion for Unified Visual Grounding (AAAI 2024 · better on 6 of 6 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- TransVG — TransVG: End-to-End Visual Grounding With Transformers (ICCV 2021) · better on 4 of 5 shared results
- MDETR — MDETR - Modulated Detection for End-to-End Multi-Modal Understanding (ICCV 2021) · better on 2 of 2 shared results
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
- Referencing Where to Focus: Improving Visual Grounding with Referential Query (NeurIPS 2024 · similarity 0.91)
- Connecting the Dots: Training-Free Visual Grounding via Agentic Reasoning (AAAI 2026 · similarity 0.91)
- Your Large Vision-Language Model Only Needs A Few Attention Heads For Visual Grounding (CVPR 2025 · similarity 0.90)
- Improved Visual Grounding through Self-Consistent Explanations (CVPR 2024 · similarity 0.90)
- SimVG: A Simple Framework for Visual Grounding with Decoupled Multi-modal Fusion (NeurIPS 2024 · similarity 0.89)
- When Visual Grounding Meets Gigapixel-level Large-scale Scenes: Benchmark and Approach (CVPR 2024 · similarity 0.89)
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.
