Research map / Vision-language and multimodal / Human and embodied modeling
Visual Abductive Reasoning
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 |
|---|---|---|---|
| Use Reasoner instead of PDVC | VAR · VAR test, premise events | BERT-S: 29.37 -> 34.88 | Helped |
| Use Reasoner instead of PDVC | VAR · VAR test, explanation events | BERT-S: 27.80 -> 30.64 | Helped |
| Use Reasoner instead of PDVC | ActivityNet Captions · ae-val | CIDEr: +2.81 | Helped |
| Add one refinement decoder after the initial decoder (ablation) | Reasoner · VAR test, all events averaged | CIDEr: 32.72 -> 36.13 | Helped |
| Add confidence embedding to cascaded reasoning (ablation) | Reasoner · VAR test, all events averaged | CIDEr: 35.22 -> 36.13 | Helped |
| Use contextualized directional position embedding instead of absolute position embedding (ablation) | Reasoner · VAR test | BLEU@4, CIDEr, and BERT-S | Helped |
| Use contextualized directional position embedding instead of context-free directional relative position embedding (ablation) | Reasoner · VAR test | BLEU@4, CIDEr, and BERT-S | Helped |
| Add hypothesis reconstruction loss (ablation) | Reasoner · VAR test | BLEU@4, CIDEr, and BERT-S | Helped |
Closest papers by meaning
- AbductiveMLLM: Boosting Visual Abductive Reasoning Within MLLMs (AAAI 2026 · similarity 0.95)
- Black Swan: Abductive and Defeasible Video Reasoning in Unpredictable Events (CVPR 2025 · similarity 0.90)
- NL-Eye: Abductive NLI For Images (ICLR 2025 · similarity 0.89)
- VisRes Bench: On Evaluating the Visual Reasoning Capabilities of VLMs (CVPR 2026 · similarity 0.87)
- Ambiguity-Aware Abductive Learning (ICML 2024 · similarity 0.86)
- Multi-step Visual Reasoning with Visual Tokens Scaling and Verification (NeurIPS 2025 · similarity 0.86)
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.