Two-Stream Region Convolutional 3D Network for Temporal Activity Detection

5 Jun 2019Huijuan XuAbir DasKate Saenko

We address the problem of temporal activity detection in continuous, untrimmed video streams. This is a difficult task that requires extracting meaningful spatio-temporal features to capture activities, accurately localizing the start and end times of each activity... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Action Recognition THUMOS’14 Two-stream R-C3D (Sum) + OHEM [email protected] 56.9 # 3
[email protected] 54.7 # 3
[email protected] 51.2 # 4
[email protected] 43.0 # 4
[email protected] 36.1 # 3
Action Recognition THUMOS’14 Single-stream R-C3D + OHEM [email protected] 57.4 # 2
[email protected] 54.9 # 2
[email protected] 51.1 # 5
[email protected] 43.1 # 3
[email protected] 35.8 # 4

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet