Research map / Vision-language and multimodal / Human and embodied modeling
Counterfactual Contrastive Learning for Weakly-Supervised Vision-Language Grounding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 counterfactual contrastive learning (CCL) to MIL-based VGN | VGN · Charades-STA | R@1 IoU=0.5: 30.77 -> 33.21 | Helped |
| add CCL to VGN | VGN · ActivityCaption | R@1 IoU=0.3: 46.17 -> 50.12 | Helped |
| combine all three transformations instead of feature-level only (ablation) | VGN+CCL · Charades-STA / ActivityCaption | R@1: 32.06 -> 33.21; 50.53 -> 50.12 | Mixed |
| add ranking loss (ablation) | VGN+CCL · Charades-STA | R@1 IoU=0.5: 31.16 -> 33.21 | Helped |
| add consistency loss (ablation) | VGN+CCL · Charades-STA | R@1 IoU=0.5: 32.55 -> 33.21 | Helped |
| increase transformations J from 3 to 6 (ablation) | VGN+CCL · Charades-STA | R@1 IoU=0.5: 33.21 -> 33.65 | No clear effect |
| CCL instead of hard negative mining | VGN · Charades-STA | R@1 IoU=0.5: 31.19 -> 33.21 | Helped |
Closest papers by meaning
- Scaling Test-Time Robustness of Vision-Language Models via Self-Critical Inference Framework (CVPR 2026 · similarity 0.86)
- FFF: Fixing Flawed Foundations in Contrastive Pre-Training Results in Very Strong Vision-Language Models (CVPR 2024 · similarity 0.86)
- V-CECE: Visual Counterfactual Explanations via Conceptual Edits (NeurIPS 2025 · similarity 0.86)
- Multi-Scale Contrastive Learning for Video Temporal Grounding (AAAI 2025 · similarity 0.85)
- CF-VLM：CounterFactual Vision-Language Fine-tuning (NeurIPS 2025 · similarity 0.85)
- SimVG: A Simple Framework for Visual Grounding with Decoupled Multi-modal Fusion (NeurIPS 2024 · similarity 0.85)
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.