Enhancement algorithm for generative digital art based on pulse-coupled neural networks

With the increasing integration of artificial intelligence into creative fields, digital art has emerged as a prominent medium for innovation. This study aims to enhance the visual quality and artistic expressiveness of generative digital artworks using a biologically inspired neural model—pulse-coupled neural networks (PCNNs). The proposed enhancement algorithm builds upon the traditional PCNN framework by incorporating an adaptive enhancement mechanism that accounts for image structure and human visual perception. Key improvements include dynamic adjustment of coupling strength and firing thresholds among neurons to refine artistic features such as texture, contrast, and color richness. Iterative optimization and adaptive feedback further support real-time enhancement during the generation process. A large dataset comprising Van Gogh-style paintings and modern abstract digital artworks was used to evaluate performance. Experimental results demonstrate that the enhanced PCNN algorithm significantly improves both the clarity and esthetic appeal of generated images. Compared to standard generative adversarial networks (GANs), the proposed method achieves approximately a 15% increase in visual perception metrics and delivers superior results in texture refinement and artistic coherence. Additionally, computational efficiency is optimized to enable real-time generation without compromising quality. This study introduces an innovative approach that bridges neuroscience-inspired models and digital art generation. By leveraging PCNNs’ dynamic spiking behavior and perceptual alignment, the algorithm provides a novel pathway for producing high-quality, visually compelling digital artworks, offering practical applications in creative AI and digital media design.