Facets of Visual Cognition in Bees: Unlocking Nature’s Blueprint for Efficient Object Recognition

Object recognition is a fundamental cognitive task, central to how biological organisms interact with their environments. This capacity enables identification, categorisation, and informed decision-making, forming the basis of perception, navigation, and goaldirected behaviour. In humans and many animals, object recognition is not only fast but also remarkably robust against environmental variability, such as changes in lighting, orientation, or scale. Yet, despite advances in artificial intelligence (AI), replicating this capacity in artificial systems remains a significant challenge. One of the major limitations in AI lies in our incomplete understanding of how biological systems, particularly those with compact and efficient neural architectures, achieve these feats of visual cognition. Among biological systems, bees represent a compelling model. Despite their miniature brains and limited number of neurons (~1 million, compared to ~86 billion in humans), bees demonstrate complex cognitive behaviours, including high-speed visual processing, precise spatial navigation, and object (flower) recognition under varying conditions. Bees achieve this with exceptional metabolic and computational efficiency – a feature that aligns with key challenges in robotic and AI systems constrained by power, size, and real-time processing requirements. This FACETS project proposes to explore the mechanisms of object recognition in bees through the lens of embodied cognition – the theory that cognition arises from dynamic interactions between the brain, the body, and the environment. The project leverages my multidisciplinary background in behavioural ecology, insect neuroscience, morphometrics, and computational modelling, as well as the Institut des Sciences du Mouvement (ISM)’s expertise in advanced modelling and bio-inspired technologies.