The ReLER (Recognition, LEarning, Reasoning) Lab is affiliated with AAII, University of Technology Sydney. The lab is committed to enable machines to accurately recognize the environment, adaptively understand the human interactions, and autonomously analyze the behavior through reasoning. To this end, we work on learning algorithms, computer vision, natural language, and their intersections. Concretely, we aim at developing novel methods for object, action, and event recognition, localizing the positions, segmenting the instances in images and videos, understanding documents and dialogues. Additionally, we also study model acceleration algorithms to speed up the recognition process... In the real-world scenarios, data hungry methods can probably fail due to fewer labeled data. We study weakly supervised learning, unsupervised learning, zero-shot learning, few-shot learning to adapt the model in this circumstance. Besides recognition, it is also essential for the machines to..