Research map / Vision-language and multimodal / Human and embodied modeling
Decoding Visual Neural Representations by Multimodal Learning of Brain-Visual-Linguistic Features
Changde Du, Kaicheng Fu, Jinpeng Li, Huiguang He
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 textual features to visual features | BraVL · DIR-Wiki+GOD-Wiki avg | top-5 accuracy (%): 55.51 -> 60.33 | Improved |
| add textual features to visual features | BraVL · DIR-Wiki+GOD-Wiki avg | top-1 accuracy (%): 20.08 -> 22.43 | Improved |
| use BraVL instead of MoPoE-VAE | multimodal VAE decoding · DIR-Wiki+GOD-Wiki avg | top-5 accuracy (%): 58.70 -> 60.33 | Improved |
| use Wiki articles instead of class names as text input (ablation) | BraVL (ALBERT) · GOD-Wiki | top-5 accuracy (%): 45.47 -> 57.04 | Improved |
| use Wiki articles instead of class names as text input (ablation) | BraVL (ALBERT) · DIR-Wiki | top-5 accuracy (%): 70.58 -> 74.69 | Improved |
| use Wiki articles instead of class names as text input (ablation) | BraVL (GPT-Neo) · GOD-Wiki | top-5 accuracy (%): 36.69 -> 53.54 | Improved |
| add intra-modality MI term L_intra (ablation) | BraVL · GOD-Wiki | top-5 accuracy (%): 51.44 -> 53.54 | Improved |
| add inter-modality MI term L_inter (ablation) | BraVL · GOD-Wiki | top-5 accuracy (%): 51.00 -> 53.54 | Improved |
| add both MI terms to the ELBO (ablation) | BraVL · DIR-Wiki | top-5 accuracy (%): 69.83 -> 72.39 | Improved |
| use MoPoE instead of PoE joint posterior (ablation) | BraVL · GOD-Wiki | top-5 accuracy (%): 39.56 -> 53.54 | Improved |
| use MoPoE instead of MoE joint posterior (ablation) | BraVL · DIR-Wiki | top-5 accuracy (%): 72.33 -> 72.39 | No clear effect |
| increase latent dimension beyond 32 | BraVL · DIR-Wiki+GOD-Wiki | accuracy | Hurt |
Closest papers by meaning
- $i$MIND: Insightful Multi-subject Invariant Neural Decoding (NeurIPS 2025 · similarity 0.90)
- Neuro-Vision to Language: Enhancing Brain Recording-based Visual Reconstruction and Language Interaction (NeurIPS 2024 · similarity 0.90)
- Multigranular Evaluation for Brain Visual Decoding (AAAI 2026 · similarity 0.90)
- Brain decoding: toward real-time reconstruction of visual perception (ICLR 2024 · similarity 0.89)
- Revealing Vision-Language Integration in the Brain with Multimodal Networks (ICML 2024 · similarity 0.89)
- CLIP-MUSED: CLIP-Guided Multi-Subject Visual Neural Information Semantic Decoding (ICLR 2024 · similarity 0.89)
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.
