His research interests currently lies in the intersection of computer vision, deep learning and human-computer interaction, especially domain adaptation, reid and vision&language topics. He has experiences in image classification, visual search, image captioning and reid.