Integrating Traditional Theories and Neurosymbolic Ai: Bridging Perceptual and Conceptual Processes in Object Recognition

The objective of this research proposal is to develop a neurosymbolic framework for object recognition that combines biologically inspired neural networks with the interpretability and abstraction capabilities of symbolic reasoning. This innovative approach addresses critical challenges such as viewpoint, scale, and occlusion invariances, bridging the gap between computational models and cognitive strategies observed in humans and non-human primates. By integrating top-down feedback mechanisms and structural analysis modules, the research aims to advance our understanding of object recognition while developing more robust, interpretable AI systems that are aligned with cognitive principles. The potential impacts of this project are far-reaching: For society, the research fosters the development of AI technologies with applications in healthcare, assistive systems, and public safety, enhancing quality of life through improved transparency and reliability. Economically, it supports industries which rely on advanced pattern recognition applications, including robotics, automation, and human-computer interaction, driving innovation and strengthening international economic competitiveness in AI-driven sectors. Academically, the project contributes to interdisciplinary insights, producing high-impact publications, creating collaboration between neuroscience, psychology, and artificial intelligence, and providing training opportunities for the next generation of researchers. By aligning computational approaches with biological principles, this research sets a foundation for transformative advancements across multiple domains in science, industry, and society.