My research focus covers deep learning in general, with specific interests in model compression/efficiency, deep generative model, and adversarial learning. Currently, I focus on productionizing these techniques to solve industry problems in computer vision, natural language processing, and multimodal learning.