Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Long-term Temporal Convolutions for Action Recognition

About

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%).

G\"ul Varol, Ivan Laptev, Cordelia Schmid• 2016

Related benchmarks

TaskDatasetResultRank
Action RecognitionUCF101
Accuracy91.7
365
Action RecognitionUCF101 (mean of 3 splits)
Accuracy92.7
357
Action RecognitionUCF101 (test)
Accuracy91.7
307
Action RecognitionHMDB51 (test)
Accuracy0.648
249
Action RecognitionHMDB-51 (average of three splits)
Top-1 Acc48.7
204
Action RecognitionHMDB51
3-Fold Accuracy67.2
191
Action RecognitionUCF101 (3 splits)
Accuracy91.7
155
Action ClassificationHMDB51 (over all three splits)
Accuracy64.8
121
Video Action RecognitionHMDB-51 (3 splits)
Accuracy64.8
116
Video Action RecognitionHMDB51 (avg over all splits)
Top-1 Acc64.8
56
Showing 10 of 12 rows

Other info

Code

Follow for update