Beyond short snippets: Deep networks for video classification

• Published in: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) pp. 4694 - 4702
• Main Authors: Ng, Joe Yue-Hei, Hausknecht, Matthew, Vijayanarasimhan, Sudheendra, Vinyals, Oriol, Monga, Rajat, Toderici, George
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.06.2015
• Subjects: Computer architecture; Computer vision; Image detection; Image recognition; Logic gates; Networks; Neural networks; Object recognition; Optical imaging; Pattern recognition; Retrieval; Segmentation; Time-domain analysis; Training
• ISSN: 1063-6919; 1063-6919; 2575-7075
• DOI: 10.1109/CVPR.2015.7299101

Convolutional neural networks (CNNs) have been extensively applied for image recognition problems giving state-of-the-art results on recognition, detection, segmentation and retrieval. In this work we propose and evaluate several deep neural network architectures to combine image information across a video over longer time periods than previously attempted. We propose two methods capable of handling full length videos. The first method explores various convolutional temporal feature pooling architectures, examining the various design choices which need to be made when adapting a CNN for this task. The second proposed method explicitly models the video as an ordered sequence of frames. For this purpose we employ a recurrent neural network that uses Long Short-Term Memory (LSTM) cells which are connected to the output of the underlying CNN. Our best networks exhibit significant performance improvements over previously published results on the Sports 1 million dataset (73.1% vs. 60.9%) and the UCF-101 datasets with (88.6% vs. 88.0%) and without additional optical flow information (82.6% vs. 73.0%).