Research map / Vision-language and multimodal / Human and embodied modeling
Open-Domain, Content-Based, Multi-Modal Fact-Checking of Out-of-Context Images via Online Resources
Check 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 web evidence and multimodal consistency reasoning | CLIP image-caption baseline · NewsCLIPpings test set | accuracy: 66.1% -> 84.7% | Helped |
| Add visual evidence images (ablation) | CCN · NewsCLIPpings test set | accuracy: 62.5% -> 84.7% | Helped |
| Add textual evidence captions (ablation) | CCN · NewsCLIPpings test set | accuracy: 57.4% -> 84.7% | Helped |
| Add evidence entities (ablation) | CCN · NewsCLIPpings test set | accuracy: 71.8% -> 84.7% | Helped |
| Add batch normalization after each component (ablation) | CCN · NewsCLIPpings test set | accuracy: 73.5% -> 84.2% | Helped |
| Use separate modality memories rather than a unified memory (ablation) | CCN · NewsCLIPpings test set | accuracy: 81.7% -> 84.7% | Helped |
| Filter exact evidence matches from the same website (ablation) | CCN · NewsCLIPpings test set | accuracy: 84.7% -> 80.3% | No clear effect |
| Add the CLIP image-caption component (ablation) | CCN · NewsCLIPpings test set | accuracy: 82.6% -> 84.7% | Helped |
| Use BERT+LSTM instead of a sentence transformer (ablation) | CCN · NewsCLIPpings test set | accuracy: 83.9% -> 84.7% | Helped |
| Add evidence-domain embeddings (ablation) | CCN · NewsCLIPpings test set | accuracy: 83.9% -> 84.7% | Helped |
| Add Places365 scene representations (ablation) | CCN · NewsCLIPpings test set | accuracy: 83.4% -> 84.7% | Helped |
| Add overlapping image-label features (ablation) | CCN · NewsCLIPpings test set | accuracy: 83.9% -> 84.7% | Helped |
Closest papers by meaning
- SNIFFER: Multimodal Large Language Model for Explainable Out-of-Context Misinformation Detection (CVPR 2024 · similarity 0.87)
- Pioneering Explainable Video Fact-Checking with a New Dataset and Multi-role Multimodal Model Approach (AAAI 2025 · similarity 0.86)
- DEFAME: Dynamic Evidence-based FAct-checking with Multimodal Experts (ICML 2025 · similarity 0.86)
- AVerImaTeC: A Dataset for Automatic Verification of Image-Text Claims with Evidence from the Web (NeurIPS 2025 · similarity 0.86)
- Beyond Text: Fine-Grained Multi-Modal Fact Verification with Hypergraph Transformers (AAAI 2025 · similarity 0.86)
- Unleashing the Potential of Consistency Learning for Detecting and Grounding Multi-Modal Media Manipulation (CVPR 2025 · 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.