Research map / Vision-language and multimodal / Human and embodied modeling
Bridging the Gap between Brain and Machine in Interpreting Visual Semantics: Towards Self-adaptive Brain-to-Text Decoding
Jiaxuan Chen, Yu Qi, Yueming Wang, Gang Pan
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 self-adaptive semantic decoding (SASD) (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | METEOR: 20.09 -> 37.11 | Improved |
| add SASD (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | BLEU-4: 14.78 -> 19.23 | Improved |
| add SASD (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | CIDEr: 39.14 -> 51.43 | Improved |
| add SASD (ablation) | Mind-SA-Small · NSD brain-to-text (subject 1) | METEOR: 19.99 -> 37.17 | Improved |
| add SASD (ablation) | Mind-SA-Large · NSD brain-to-text (subject 1) | METEOR: 20.48 -> 36.68 | Improved |
| increase fMRI encoder from 8 to 24 layers (with SASD) (ablation) | Mind-SA · NSD brain-to-text (subject 1) | CIDEr: 50.13 -> 47.94 | No clear effect |
| increase fMRI encoder from 8 to 16 layers (with SASD) (ablation) | Mind-SA · NSD brain-to-text (subject 1) | CIDEr: 50.13 -> 51.43 | No clear effect |
| use 16 captured patches instead of 4 (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | METEOR: 20.47 -> 37.11 | Improved |
| use 16 captured patches instead of 8 (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | METEOR: 32.34 -> 37.11 | Improved |
| use 24 captured patches instead of 16 (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | METEOR: 37.11 -> 36.50 | Mixed |
| use 24 captured patches instead of 16 (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | BLEU-4: 19.23 -> 19.34 | Mixed |
| use 24 captured patches instead of 16 (ablation) | Mind-SA-Base · NSD brain-to-text (subject 1) | CIDEr: 51.43 -> 50.67 | No clear effect |
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.
- UMBRAE — UMBRAE: Unified Multimodal Brain Decoding (ECCV 2024) · 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
- $i$MIND: Insightful Multi-subject Invariant Neural Decoding (NeurIPS 2025 · similarity 0.90)
- MindAlign: Bridging EEG, Vision, and Language for Zero-Shot Visual Decoding (NeurIPS 2026 · similarity 0.90)
- Neuro-Vision to Language: Enhancing Brain Recording-based Visual Reconstruction and Language Interaction (NeurIPS 2024 · similarity 0.89)
- Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain) (ICLR 2025 · similarity 0.88)
- Finding Shared Decodable Concepts and their Negations in the Brain (ICLR 2025 · similarity 0.88)
- Multigranular Evaluation for Brain Visual Decoding (AAAI 2026 · 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.
