Biography
I am doing my PhD at HHMI Janelia Research Campus as part of the joint graduate program with Johns Hopkins university. My current work focuses on using a brain-machine interface to study the role of behavioral and sensory signals in the mouse visual cortex.
My previous work involved developing machine learning tools for studying large-scale neuroscience datasets. I developed Facemap for predicting activity of ~50,000 neurons recorded with 2-photon calcium imaging in mice. Facemap predicted activity driven by mouse orofacial movements across different cortical regions. It includes a pose estimation model for tracking distinct keypoints on the mouse face and a neural network model for predicting neural activity using the pose estimates. I also developed a segmentation tool using the Facemap model for tracking soft-body objects like the mouse tongue. Facemap provides state-of-the-art performance for tracking mouse orofacial movements with a processing speed several times faster than existing models, making it a powerful tool for real-time experimental interventions.
Projects
- Facemap: pose tracking and neural activity prediction model
- Segmentation model for soft-body object (tongue)
- Rastermap (GUI)
Research interests
- Computational neuroscience
- Deep learning