Daily View
Thursday, April 19, 2018

Deep Learning for Computational Video Understanding (Thu. 19 Apr, 2018 10:00 am)

Over the last few years, deep learning has revolutionized the field of still-image analysis by delivering breakthrough results on many hard computer vision tasks, including object recognition, detection, scene classification, and semantic pixel-level prediction. While there has been a widespread expectation that these performance improvements should naturally extend to the video domain, the results so far have been lagging compared to the image setting.   In this talk, I will discuss the unique challenges posed by the video domain and provide a survey of recent efforts in designing effective deep computational models for video understanding. I will conclude by presenting a deep model for spatiotemporal visual attention explicitly trained to mimic where humans look in a video and I will describe how it can be leveraged to improve the performance of existing algorithms for action recognition.