Research map / Vision-language and multimodal / Human and embodied modeling
Webly Supervised Knowledge Embedding Model for Visual Reasoning
Wenbo Zheng, Lan Yan, Chao Gou, Fei-Yue Wang
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 web-image and noisy-tag training in Stage II (ablation) | Ours · CLEVR | Overall accuracy: 99.3 -> 99.8 | Improved |
| add web-image and noisy-tag training in Stage II (ablation) | Ours · NLVR Dev | Accuracy: 72.4 -> 81.3 | Improved |
| add web-image and noisy-tag training in Stage II (ablation) | Ours · NLVR Test-P | Accuracy: 74.3 -> 80.6 | Improved |
| add language modulation and the key-value memory network (ablation) | Ours · NLVR Dev | Accuracy: 60.4 -> 81.3 | Improved |
| add language modulation and the key-value memory network (ablation) | Ours · NLVR Test-P | Accuracy: 62.3 -> 80.6 | Improved |
| add the key-value memory network (ablation) | Ours · NLVR Dev | Accuracy: 68.1 -> 81.3 | Improved |
| add the key-value memory network (ablation) | Ours · NLVR Test-P | Accuracy: 70.1 -> 80.6 | Improved |
| add language modulation (ablation) | Ours · NLVR Dev | Accuracy: 70.5 -> 81.3 | Improved |
| add language modulation (ablation) | Ours · NLVR Test-P | Accuracy: 70.4 -> 80.6 | Improved |
| add the key-value memory network (ablation) | Ours · CLEVR | Overall accuracy: 97.0 -> 99.8 | Improved |
| add language modulation (ablation) | Ours · CLEVR | Overall accuracy: 99.0 -> 99.8 | Improved |
| use the proposed webly supervised model instead of CAN | CLEVR · CLEVR | Overall accuracy: 98.9 -> 99.8 | Improved |
Closest papers by meaning
- Semantic-Enriched Latent Visual Reasoning (ICML 2026 · similarity 0.89)
- Multi-step Visual Reasoning with Visual Tokens Scaling and Verification (NeurIPS 2025 · similarity 0.88)
- Monet: Reasoning in Latent Visual Space Beyond Image and Language (CVPR 2026 · similarity 0.88)
- SLVR: Structured Latent Visual Reasoning via Human-like Reasoning Flows (NeurIPS 2026 · similarity 0.88)
- UKnow: A Unified Knowledge Protocol with Multimodal Knowledge Graph Datasets for Reasoning and Vision-Language Pre-Training (NeurIPS 2024 · similarity 0.88)
- MMMG: A Massive, Multidisciplinary, Multi-Tier Generation Benchmark for Text-to-Image Reasoning (NeurIPS 2025 · similarity 0.88)
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.
