Search
The human brain is capable of effortlessly integrating information from various sensory modalities, such as vision and hearing. Leveraging neural integration principles—including superadditivity, inverse effectiveness, and temporal coherence—it achieves highly efficient multimodal perception fusion with remarkably low energy consumption. In contrast, most contemporary artificial intelligence systems rely on conventional algorithmic frameworks that follow a sequential "sense-then-fuse" strategy. These systems lack the adaptive and nonlinear coupling characteristics inherent in biological neural networks. As a result, deploying AI systems on edge hardware presents significant challenges in terms of energy efficiency and processing latency, ultimately constraining the performance of edge-based intelligent perception.
-
C114 Communication Network -
Communication Home
