Rethinking Motion Representation: Residual Frames with 3D ConvNets for Better Action Recognition
This work addresses the efficiency problem in video action recognition for researchers and practitioners by offering a faster alternative to optical flow, though it is incremental as it builds on existing 3D ConvNet methods.
The paper tackles the high computational cost of optical flow in action recognition by proposing residual frames as input to 3D ConvNets, achieving improvements of 20.5% and 12.5% in top-1 accuracy on UCF101 and HMDB51 datasets when trained from scratch.
Recently, 3D convolutional networks yield good performance in action recognition. However, optical flow stream is still needed to ensure better performance, the cost of which is very high. In this paper, we propose a fast but effective way to extract motion features from videos utilizing residual frames as the input data in 3D ConvNets. By replacing traditional stacked RGB frames with residual ones, 20.5% and 12.5% points improvements over top-1 accuracy can be achieved on the UCF101 and HMDB51 datasets when trained from scratch. Because residual frames contain little information of object appearance, we further use a 2D convolutional network to extract appearance features and combine them with the results from residual frames to form a two-path solution. In three benchmark datasets, our two-path solution achieved better or comparable performances than those using additional optical flow methods, especially outperformed the state-of-the-art models on Mini-kinetics dataset. Further analysis indicates that better motion features can be extracted using residual frames with 3D ConvNets, and our residual-frame-input path is a good supplement for existing RGB-frame-input models.