Synergistic Optimization Mechanism of Brain-Inspired Intelligence and Generative Models in Small Target Detection

In the era of generative artificial intelligence, we attempt to apply brain-inspired intelligence and generative models to small target detection, but currently lack a synergistic optimization mechanism. This paper attempts to enhance the detection performance of small targets by integrating the bionic perception of brain-inspired intelligence with the data augmentation capabilities of generative models. Firstly, we simulate the multi-scale attention mechanism of the biological visual system using a brain-inspired spiking neural network (SNN), enhancing the spatial sensitivity of small target detection. Secondly, we generate high-quality small target samples and optimize local feature representation through Generative Adversarial Networks (GANs), alleviating data scarcity and feature ambiguity. This paper proposes for the first time a solution providing a cross-modal collaborative framework for small target detection, exploring the potential of combining brain-inspired mechanisms with generative models.