Research map / Vision-language and multimodal / Human and embodied modeling
Confidence-aware Pseudo-label Learning for Weakly Supervised Visual Grounding
Yang Liu, Jiahua Zhang, Qingchao Chen, Yuxin Peng
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 Heuristic+ improvements over Pseudo-Q heuristic pipeline (ablation) | CPL · RefCOCO+ val | top-1 accuracy (%): 38.88 -> 40.36 | Improved |
| add uni-modal real query propagation (ablation) | CPL with Heuristic+ · RefCOCO+ val | top-1 accuracy (%): 40.36 -> 50.65 | Improved |
| add cross-modal verification module and selective grounding loss (ablation) | CPL · RefCOCO+ val | top-1 accuracy (%): 51.08 -> 51.81 | Improved |
| add object-centric and relation-aware BLIP pipelines on top of H+ with real-query (ablation) | CPL · RefCOCO+ val and testB | top-1 accuracy (%): val 50.65 -> 51.08; testB 45.52 -> 44.46 | Mixed |
| use BLIP alone to select proposals (ReBLIP) (ablation) | visual grounding training · RefCOCO+ val | top-1 accuracy (%): 12.07 vs Pseudo-Q 38.88 | Hurt |
| increase number of object proposals from 4 to 10 per pipeline (ablation) | CPL · RefCOCO+ val | top-1 accuracy (%): 40.83 -> 48.77 | Improved |
| increase pseudo-query number from 200 to all (ablation) | CPL · RefCOCO+ val | top-1 accuracy (%): 49.94 -> 51.81 | Improved |
| swap in our detectors for Pseudo-Q (ablation) | Pseudo-Q · RefCOCO+ val | top-1 accuracy (%): 38.88 -> 38.03 | No clear effect |
| fine-tune with 10% labeled data | CPL · RefCOCO+ val | top-1 accuracy (%): 50.65 -> 64.76 | Improved |
| freeze BERT during training (ablation) | CPL (uni-modal) · RefCOCO+ val | top-1 accuracy (%): 50.65 -> 46.19 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Pseudo-Q — Pseudo-Q: Generating Pseudo Language Queries for Visual Grounding (CVPR 2022) · 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
- Momentum Pseudo-Labeling for Weakly Supervised Phrase Grounding (AAAI 2025 · similarity 0.89)
- Improved Visual Grounding through Self-Consistent Explanations (CVPR 2024 · similarity 0.88)
- Referencing Where to Focus: Improving Visual Grounding with Referential Query (NeurIPS 2024 · similarity 0.87)
- Connecting the Dots: Training-Free Visual Grounding via Agentic Reasoning (AAAI 2026 · similarity 0.87)
- Look Around Before Locating: Considering Content and Structure Information for Visual Grounding (AAAI 2025 · similarity 0.87)
- Cross-Modal Match for Language Conditioned 3D Object Grounding (AAAI 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.
