Improving the efficiency of current neural networks and modeling them on biological neural systems have become prominent research directions in recent years. The pulse-coupled neural network (PCNN) is widely used to mimic the computational characteristics of the human brain in computer vision and neural network fields. However, PCNN faces limitations such as limited neural connections, high computational costs, and a lack of stochastic properties. This study proposes a random-coupled neural network (RCNN) to address these limitations. RCNN employs a stochastic inactivation process, selectively inactivating neural connections using a random inactivation weight matrix. This method reduces the computational burden and allows for extensive neural connections. RCNN encodes constant stimuli as periodic spike trains and periodic stimuli as chaotic spike trains, reflecting the information encoding characteristics of biological neural systems. Our experiments applied RCNN to image segmentation and fusion tasks, demonstrating its robustness, efficiency, and high noise resistance. Results indicate that RCNN surpasses traditional methods in performance across these applications.

Precise particle picking in cryo-electron microscopy (cryo-EM) single-particle analysis constitutes a fundamental challenge in achieving high-resolution 3D reconstruction. To address low detection rates and elevated false positives caused by low signal-to-noise ratios (SNRs) and weak image contrast, we developed a novel semantic-instance collaborative picking framework. Key innovations include: (i) A multi-frequency adaptive U-Net framework that precisely localizes particles via global-context semantic modeling and multi-scale feature fusion; (ii) CryoSIP further enhances the collaborative optimization between SAM and U-Net by refining the interaction between SAM and U-Net–derived semantic priors, thereby improving instance mask generation. Experiments demonstrate that our framework reduces false positives significantly compared to state-of-the-art tools while achieving high recall in particle picking. 3D reconstructions using this approach exhibit improved density map resolution, demonstrating its potential for advancing atomic-resolution cryo-EM structural analysis.