VIBE: Video-Input Brain Encoder for fMRI Response Modeling
Schad, Daniel Carlström, Dixit, Shrey, Keck, Janis, Studenyak, Viktor, Shpilevoi, Aleksandr, Bicanski, Andrej
–arXiv.org Artificial Intelligence
We present VIBE, a two-stage Transformer that fuses multi-modal video, audio, and text features to predict fMRI activity. Representations from open-source models (Qwen2.5, BEATs, Whisper, SlowFast, V-JEPA) are merged by a modality-fusion transformer and temporally decoded by a prediction transformer with rotary embeddings. Trained on 65 hours of movie data from the CNeuroMod dataset and ensembled across 20 seeds, VIBE attains mean parcel-wise Pearson correlations of 0.3225 on in-distribution Friends S07 and 0.2125 on six out-of-distribution films. An earlier iteration of the same architecture obtained 0.3198 and 0.2096, respectively, winning Phase-1 and placing second overall in the Algonauts 2025 Challenge.
arXiv.org Artificial Intelligence
Jul-28-2025
- Country:
- Asia > Middle East
- Jordan (0.04)
- Saudi Arabia > Asir Province
- Abha (0.04)
- Europe > Germany
- Asia > Middle East
- Genre:
- Research Report (0.84)
- Industry:
- Health & Medicine > Therapeutic Area
- Neurology (1.00)
- Leisure & Entertainment (0.93)
- Media > Television (0.68)
- Health & Medicine > Therapeutic Area
- Technology: