Mind-Omni: A Unified Multi-Task Framework for Brain-Vision-Language Modeling via Discrete Diffusion

Researchers introduce Mind-Omni, a unified framework that consolidates seven brain-computer interface tasks through discrete diffusion modeling, using a novel Brain Tokenizer to convert continuous neural signals into standardized tokens. The multi-task approach demonstrates competitive or superior performance compared to specialized models while enabling cross-modal interactions between brain, vision, and language data.

Mind-Omni represents a significant architectural shift in brain-computer interface research by moving away from task-specific models toward a generalized framework capable of handling multiple encoding and decoding operations simultaneously. The innovation centers on the Brain Tokenizer, which standardizes heterogeneous neural signal data into discrete tokens—a critical engineering challenge that enables seamless interaction across different data modalities within a shared semantic space. This approach mirrors broader trends in AI toward foundation models that leverage multi-task learning to improve generalization and efficiency.

The research community has historically fragmented around specialized models for each BCI application, limiting knowledge transfer and requiring substantial retraining for new tasks. Mind-Omni's unified approach addresses this inefficiency by demonstrating that multi-task synergy can deliver performance competitive with or exceeding larger, task-specific models. The inclusion of a Brain Question Answering instruction-tuning dataset adds reasoning capabilities beyond traditional encoding-decoding paradigms, positioning the framework closer to true multimodal understanding.

For the broader AI and neuroscience sectors, this work validates that neural signals can be modeled within standardized semantic spaces—a prerequisite for scaling BCI applications across clinical, research, and consumer domains. The competitive performance against specialized models suggests the framework could accelerate BCI commercialization by reducing development costs and training time. Investors and researchers should monitor whether this architecture becomes the standard for future BCI development, as framework dominance typically creates significant competitive advantages.

• →Mind-Omni unifies seven distinct BCI tasks through discrete diffusion, eliminating the need for separate specialized models.
• →A novel Brain Tokenizer converts continuous neural signals into standardized discrete tokens enabling cross-modal understanding.
• →The framework achieves competitive or superior performance versus larger single-task models, validating multi-task synergy benefits.
• →Addition of Brain Question Answering dataset enables advanced reasoning beyond traditional encoding-decoding operations.
• →Open-source release positions the framework as a potential standard for future BCI and neural modeling research.