Activity Recognition based on a Magnitude-Orientation Stream Network
This work addresses activity recognition for video analysis, but it is incremental as it builds on existing two-stream methods with a modified temporal input.
The paper tackled activity recognition in videos by proposing a Magnitude-Orientation Stream (MOS) to enhance motion learning in two-stream convolutional networks, resulting in improved performance on HMDB51 and UCF101 datasets.
The temporal component of videos provides an important clue for activity recognition, as a number of activities can be reliably recognized based on the motion information. In view of that, this work proposes a novel temporal stream for two-stream convolutional networks based on images computed from the optical flow magnitude and orientation, named Magnitude-Orientation Stream (MOS), to learn the motion in a better and richer manner. Our method applies simple nonlinear transformations on the vertical and horizontal components of the optical flow to generate input images for the temporal stream. Experimental results, carried on two well-known datasets (HMDB51 and UCF101), demonstrate that using our proposed temporal stream as input to existing neural network architectures can improve their performance for activity recognition. Results demonstrate that our temporal stream provides complementary information able to improve the classical two-stream methods, indicating the suitability of our approach to be used as a temporal video representation.