任俊博士学术报告通知

报告题目：基于深度表示学习的行为识别

时间：2018年1月8日16:30-18:00

地点：6C702

讲座介绍：

The Attention-focusing mechanism in the human brain plays an important role when humans attempt to recognize movements or actions, in the presence of objects in the scene. This presentation will introduce how to enable the computer to gain a high-level understanding of what people are doing in front of cameras, mimicking the attention-focusing mechanisms of the human visual system. First, he will introduce our latest proposed CNN model, trained with new objective function, which can help to train a robust model for action recognition. Then, he will talk about the novel RNN model, which can be easily transferred into other time-series problems. Lastly, he shares some toolkits and scripts that can help researchers to improve the efficiency of writing papers/reports.

报告人介绍：

任俊， 2014年毕业于成都理工大学，获硕士学位。受国家留学基金委资助，在新西兰梅西大学自然与数理学院计算机专业攻读博士学位。研究领域包括机器学习，视频处理，人工智能等。任俊博士目前已在计算机科学领域发表十余篇高质量论文。

网络安全学院

2018年1月5日