Search papers, labs, and topics across Lattice.
This study introduces Brain-Inspired Capture (BI-Cap), a neuromimetic perceptual simulation paradigm that enhances visual decoding in brain-computer interfaces by emulating the processing mechanisms of the Human Visual System. By integrating biologically plausible transformations and employing Mutual Information-guided dynamic blur regulation, BI-Cap effectively aligns neural and visual modalities while addressing the non-stationarity of neural activity through an evidence-driven latent space representation. The approach demonstrates significant performance improvements over existing methods, achieving relative gains of 9.2% and 8.0% in zero-shot brain-to-image retrieval tasks across two public benchmarks.
BI-Cap achieves a remarkable 9.2% improvement in brain-to-image retrieval by mimicking the Human Visual System's processing, bridging the gap between neural signals and visual outputs.
Visual decoding of neurophysiological signals is a critical challenge for brain-computer interfaces (BCIs) and computational neuroscience. However, current approaches are often constrained by the systematic and stochastic gaps between neural and visual modalities, largely neglecting the intrinsic computational mechanisms of the Human Visual System (HVS). To address this, we propose Brain-Inspired Capture (BI-Cap), a neuromimetic perceptual simulation paradigm that aligns these modalities by emulating HVS processing. Specifically, we construct a neuromimetic pipeline comprising four biologically plausible dynamic and static transformations, coupled with Mutual Information (MI)-guided dynamic blur regulation to simulate adaptive visual processing. Furthermore, to mitigate the inherent non-stationarity of neural activity, we introduce an evidence-driven latent space representation. This formulation explicitly models uncertainty, thereby ensuring robust neural embeddings. Extensive evaluations on zero-shot brain-to-image retrieval across two public benchmarks demonstrate that BI-Cap substantially outperforms state-of-the-art methods, achieving relative gains of 9.2\% and 8.0\%, respectively. We have released the source code on GitHub through the link https://github.com/flysnow1024/BI-Cap.
