Research map / Vision-language and multimodal / Human and embodied modeling
AVID: A 5T fMRI Dataset for Benchmarking Auditory-induced Visual Mental Imagery Decoding
Shiqi Shen, Shurui Li, Yuanning Li, Xilin Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Introduces AVID, a 5T fMRI dataset for benchmarking decoding of auditory-induced visual mental imagery, aimed at training-scale VMI decoding research.
Problem
Visual mental imagery decoding from brain activity is limited by the lack of dedicated training-scale datasets; existing paradigms rely on visual prompts or short-term memory images.
Approach
A dataset in which visual mental imagery is induced by auditory cues rather than visual prompts or short-term memory images, recorded with 5T fMRI at training scale.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| single-participant MindEye1 versus leave-one-participant-out transfer MindEye2 | fMRI-to-image reconstruction · AVID complex split | CLIP-T: 67.83% -> 57.95% | Hurt |
| multi-participant joint training (MindBridge) versus single-participant MindEye1 | fMRI-to-image reconstruction · AVID complex split | CLIP-T: 67.83% vs 67.66% | No clear effect |
| pixel-level metrics for imagery decoding | all four baselines · AVID complex split | PixCorr: 0.034 to 0.056 | No clear effect |
| controlled simple split versus complex split | baselines · TIFA | TIFA: 32.56% -> 48.92% (LDM) | Improved |
| SSIM on sparse stimuli | MindEye1 · AVID simple split | SSIM: 0.291 -> 0.962 while PixCorr -0.028 | Mixed |
| attribute recovery above chance for color and position, near chance for identity | baselines · AVID simple split | class-balanced accuracy: identity 2.07% to 3.38% vs chance 1.05% | Mixed |
Closest papers by meaning
- NSD-Imagery: A Benchmark Dataset for Extending fMRI Vision Decoding Methods to Mental Imagery (CVPR 2025 · similarity 0.91)
- Brain decoding: toward real-time reconstruction of visual perception (ICLR 2024 · similarity 0.90)
- MindEye2: Shared-Subject Models Enable fMRI-To-Image With 1 Hour of Data (ICML 2024 · similarity 0.89)
- See Through Their Minds: Learning Transferable Brain Decoding Models from Cross-Subject fMRI (AAAI 2025 · similarity 0.88)
- Neuro-Vision to Language: Enhancing Brain Recording-based Visual Reconstruction and Language Interaction (NeurIPS 2024 · similarity 0.88)
- Visual Decoding and Reconstruction via EEG Embeddings with Guided Diffusion (NeurIPS 2024 · similarity 0.87)
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.
