My general research interest is to develop structured machine learning techniques for computer vision tasks. I mainly investigate how to exploit the human commonsense and incorporate them to develop the advanced artificial intelligence system. Recently, I focus on Deep Learning, Graph neural networks, including Graph LSTM, structure learning Human-centric Analysis, specially on human generation, image-based tryon Multi-modality Dialogue and Interactive System, such as dialogue and robotic navigation Automatic Machine Learning, including NAS and HPO.