Srinath Nizampatnam

The paper demonstrates that a simple ON-OFF classifier can achieve robust odor recognition in locusts, suggesting a generalizable computational strategy for invariant pattern recognition.

This paper uses optimization-based modeling to study how sensory networks, specifically in the olfactory system, track low-dimensional latent representations of stimuli through temporal dynamics.

The paper investigates how odor-evoked responses in locusts vary with stimulus history and shows that a linear classifier matches behavioral decoding results.