0.25
0.5
0.75
1.25
1.5
1.75
2
One Shot Similarity Metric Learning for Action Recognition
Published on Oct 17, 20117043 Views
The One-Shot-Similarity (OSS) is a framework for classifier-based similarity functions. It is based on the use of background samples and was shown to excel in tasks ranging from face recognition to do
Related categories
Chapter list
One-Shot-Similarity Metric Learning for Action Recognition00:00
Outline00:13
Metric Learning - 100:41
Metric Learning - 201:16
Metric Learning - 301:31
Metric Learning - 402:13
Metric Learning - 502:44
One-Shot-Similarity (OSS) - 103:04
One-Shot-Similarity (OSS) - 203:44
One-Shot-Similarity (OSS) - 304:12
One-Shot-Similarity (OSS) - 404:31
One-Shot-Similarity (OSS) - 504:44
One-Shot-Similarity (OSS) - 605:01
One-Shot-Similarity (OSS) - 705:43
One-Shot-Similarity Metric Learning - 106:05
One-Shot-Similarity Metric Learning - 207:11
One-Shot-Similarity Metric Learning - 307:33
One-Shot-Similarity Metric Learning - 408:05
OSSML – Objective Function08:13
OSSML for Action Recognition09:18
Action pair matching10:07
Challenges10:32
The ASLAN benchmark - Setting11:25
Baseline Tests - 112:02
Baseline Tests - 212:30
Baseline Results on the ASLAN set12:48
Pipeline - 113:20
Pipeline - 213:42
Pipeline - 314:05
Pipeline - 414:36
Pipeline - 514:49
OSSML Results14:56
OSSML Results (AUC)15:15
Same classified as Same (TP)15:30
Not-Same classified as Not-Same (TN)16:03
Same classified as Not-Same (FN)16:21
Not-Same classified as Same (FP)16:59
Summary17:35
Thank You!18:07