Artificial Intelligence (AI) is transforming perceptual science by offering novel ways to understand sensory perception and cognition. As computational power increases, so does our ability to model and simulate complex human perception processes. Self-organizing AI systems, particularly those utilizing machine learning algorithms like Self-Organizing Maps, have shown promise in mimicking complex perceptual phenomena. These advancements address fundamental questions about how neural mechanisms underlie sensory experiences and highlight ongoing debates concerning the limitations and capabilities of AI in replicating human-like sensory processing. While significant progress has been made in this field, gaps remain, particularly in integrating AI-driven models with empirical findings from neuroscience and psychology.

Self-organization, a major functional principle in the living brain, may be seen as the mechanistic foundation of all forms of adaptive learning and intelligence, from the single-synapse level to that of integrated neural networks in humans and machines (AI). The neuroscience, psychology and mathematics of self-organization deliver powerful models of knowledge generation across hierarchically organized levels of functional integration in adaptive intelligence, from sensory to cognitive learning and representation. Specificity, modular connectivity, and plasticity of neurons and networks allow function to grow without the necessity for increasing structural system complexity. The self-organization of plastic connectivity in sensory networks helps us understand how blind individuals can build visual representations of the physical world, physically expressed and materialized in the form of drawings, on the basis of self-reinforced activity-dependent learning that achieves stable cross-modal representations. The development of brain self-organization across evolution permits to establish the missing link between mental and physical worlds. The principles of self-organization clarify how physical properties are encoded, acted upon, and transformed by adaptive intelligence in humans and machines (AI). Novel solutions for robotic sensing and robot perception and cognition can arise from principles of self-organization.

This Research Topic aims to explore the potential of models based on the principles of self-organization for advancing our fundamental understanding of sensory cognition. The collection also seeks to answer questions relating to how AI systems can replicate human perceptual processes, or provide predictions and insights into the neural underpinnings of sensory experience. Among the objectives are understanding multisensory integration and the functional re-organization of sensory networks in the brain, and potential applications in sensory substitution and rehabilitation. The ethical and societal implications of using AI to simulate human perception are to be taken into account. The overarching goal here is to expand foundational knowledge in perceptual science, and to inspire innovative model and research methodologies that are likely to promote conceptual breakthroughs in the field.

Articles focusing on, but not limited to, the following themes are welcome:
• Self-organization in human perception and the sensory brain
• Self-organizing AI for modeling sensory integration, predictive coding, and cognitive representation in context
• Applications in sensory substitution and rehabilitation
• Interdisciplinary aspects linking psychology, neuroscience, and AI
• Ethical and societal impacts of simulating perception with AI

Contributions from varied disciplines are welcome, including psychology, computer science, and neuroscience. By encouraging collaborations across fields, this Research Topic is to inspire transformative advancements in perceptual science in alignment with interdisciplinary approaches and scientific innovation.

Are you interested in contributing a manuscript to this Research Topic?

AI-Driven Models Transforming Perceptual Science: Self-Organizing Intelligence for Sensory Cognition