A hallmark of intelligence is the ability to infer useful information from the environment, such as the presence of food or
threats. When navigating an environment, a single object can lead to an infinite number of different images on our retina, yet
we effortlessly unify the multiple views of a single object into a comprehensive identity, i.e., object recognition. Thus, the
temporal continuity of the retinal image and the underlying spatiotemporal structure of the world may serve as key cues in
invariant object recognition. However how the brain exploits temporal information to create meaningful perceptual
experiences remains unexplored. In this proposal, we seek to address this challenge and we aim to investigate how neural
representations of objects evolve through time for familiar views of objects and how they are temporally connected for novel
views. We further aim to evaluate how learning can use temporal information to reshape neural representations of objects.
Toward this goal, we plan to use our expertise in behavioral, systems, and computational neuroscience to develop a
framework that will allow us to understand the algorithm used by the mouse visual cortex to extract relevant features from
complex environments and guide natural behaviors. By implementing an object discrimination task within a virtual
environment, we will be able to identify the latent causes of object recognition, and with neural population recordings from
the higher visual system, we will discover how temporal information of objects is represented in the neural activity and how in
turn these representations affect the behavior of the animals. Overall, we expect this project to have an impact not only on
our understanding.