Weijian Deng is a Postdoctoral Research Fellow at the Australian National University. His research focuses on machine learning generalization, 3D computer vision, and the reliability of multimodal foundation models. He has published as first author in leading venues such as TPAMI, TIP, CVPR, ICML, NeurIPS, and TCSVT, with work spanning unsupervised model evaluation, robust vision–language models, and physically grounded 3D scene understanding. He also collaborates closely with industry partners on applied AI systems. His broader research goal is to develop reliable and self-improving models that can operate robustly in complex real-world 3D environments.