0.25
0.5
0.75
1.25
1.5
1.75
2
Improving Multi-label Learning with Missing Labels by Structured Semantic Correlations
Published on Oct 24, 20161933 Views
Multi-label learning has attracted significant interests in computer vision recently, finding applications in many vision tasks such as multiple object recognition and automatic image annotation. Asso
Related categories
Chapter list
Improving Multi-label Learning with Missing Labels by Structured Semantic Correlations00:00
Multi-label Image Recognition IS Everywhere00:15
Dog ? 00:35
Dog, chair carpet, indoor, high five, pet, funny…00:39
Multi-label Image Recognition IS Everywhere/100:55
Multi-label Image Recognition IS Everywhere/201:06
Multi-label Image Recognition IS Everywhere/301:16
Multi-label Image Recognition IS Everywhere/401:20
Multi-label learning01:25
Challenges01:50
Missing Labels Problem/102:37
Missing Labels Problem/202:56
Missing Labels Problem/303:04
Missing Labels Problem/403:14
Missing Labels Problem/504:02
Missing Labels Problem/604:03
Label – Label correlations04:07
Instance – Instance Correlations04:32
We want to formulate the problem04:43
Intuition05:00
Semantic Feature Extraction/105:42
Semantic Feature Extraction/205:56
Semantic Feature Extraction/306:19
Local Semantic Descriptor/106:34
Local Semantic Descriptor/206:55
Local Semantic Descriptor/306:57
Graph Construction07:10
Experimental Results/108:43
System Architecture09:37
Experimental Results/209:41
Experimental Results/310:04