Long-term Temporal Convolutions for Action Recognition
This work addresses the challenge of capturing long-range temporal structure in action recognition for video analysis, representing an incremental advancement over prior methods.
The paper tackled the problem of modeling human actions at their full temporal extent in video recognition by introducing long-term temporal convolutions (LTC) in CNNs, resulting in state-of-the-art accuracy improvements, such as 92.7% on UCF101 and 67.2% on HMDB51.
Typical human actions last several seconds and exhibit characteristic spatio-temporal structure. Recent methods attempt to capture this structure and learn action representations with convolutional neural networks. Such representations, however, are typically learned at the level of a few video frames failing to model actions at their full temporal extent. In this work we learn video representations using neural networks with long-term temporal convolutions (LTC). We demonstrate that LTC-CNN models with increased temporal extents improve the accuracy of action recognition. We also study the impact of different low-level representations, such as raw values of video pixels and optical flow vector fields and demonstrate the importance of high-quality optical flow estimation for learning accurate action models. We report state-of-the-art results on two challenging benchmarks for human action recognition UCF101 (92.7%) and HMDB51 (67.2%).