Computational principles of object and face recognition: bridging neural mechanisms and AI

This research line investigates how the brain recognizes objects and faces across diverse conditions, integrating primate neurophysiology, human behavior, and computational modeling. It focuses on hierarchical representations, specialized face processing, and linking neural activity to behavior. The aim is to clarify how visual pathways inspire advances in artificial intelligence and to develop computational theories that account for both neural and behavioral phenomena.

Publication List:
  • Emergence of brain-like mirror-symmetric viewpoint tuning in convolutional neural networks
    A. Farzmahdi, W. Zarco, W. Freiwald, N. Kriegeskorte, T. Golan, eLife (2024)
  • Mirror-symmetric viewpoint selectivity in deep neural networks
    A. Farzmahdi, T. Golan, R. Ebrahimpour, W. Freiwald, Computation and Systems Neuroscience (2021)
  • Mechanisms of Facial Tuning in a Brain-inspired Deep Network
    A. Farzmahdi, R. Ebrahimpour, W. Freiwald, Vision Science Society (2020)
  • A specialized face-processing model inspired by the organization of monkey face patches explains
    several face-specific phenomena observed in humans
    A. Farzmahdi, K. Rajaei, M. Ghodrati, R. Ebrahimpour, M. Khaligh-Razavi, Scientific Reports (2016)
  • Feedforward object-vision models only tolerate small image variations compared to human
    M. Ghodrati, A. Farzmahdi, K. Rajaei, M. Khaligh-Razavi, R. Ebrahimpour, Front. in Comput. Neurosci. (2014)

Variability, uncertainty, and probabilistic coding in visual cortex

This research explores how neural variability and correlations encode probabilistic inferences about natural scenes. By connecting image statistics to shared neural activity, it advances understanding of how the brain represents uncertainty and efficiently processes complex, real-world inputs.

Publication List:
  • Relating natural image statistics to patterns of response covariability in macaque primary visual cortex
    A. Farzmahdi, A. Kohn, R. Coen-Cagli, Nature Communications (2025)
  • Relating covariability in visual cortex to natural image statistics
    A. Farzmahdi, R. Coen-Cagli, Cognitive Computational Neuroscience (2022)

Task-dependent modulation of visual representations

This line examines how behavioral goals and categorization tasks shape visual processing and perception. It highlights how top-down signals modulate brain activity and influence performance during object recognition.

Publication List:
  • Task‐dependent neural representations of visual object categories
    A. Farzmahdi, F. Fallah, R. Rajimehr, R. Ebrahimpour, European Journal of Neuroscience (2021)