0.25
0.5
0.75
1.25
1.5
1.75
2
Segmentation-robust Representations, Matching, and Modeling for Sign Language Recognition
Published on Aug 24, 20113714 Views
Distinguishing true signs from transitional, extraneous movements made by the signer as s/he moves from one sign to the next is a serious hurdle in the design of continuous Sign Language recognition s
Related categories
Chapter list
Segmentation-robust Representations, Matching, and Modeling for Sign Language Recognition00:00
Beyond Single Gestures00:26
Related Work in Sign Language01:12
Outline02:43
Matching Single Signs in Clutter03:18
After motion, aspect, and size filtering….03:37
Multiple Candidates by Grouping04:10
Multiple Candidates – Method 204:35
Matching05:30
Instance-based Models: Dynamic Programming06:29
Constraints Control Combinatorics07:41
Best matched hands08:19
Recovered Hand09:06
Impact of Grouping09:21
frag-Hidden Markov Models10:23
Recognition for Sentences11:55
Movement Epenthesis Issue12:15
Handling ME and segmentation13:03
Enhanced Level Building14:21
Error rates15:54
Impact of Grammar16:49
Across Signers (Purdue Dataset)16:53
Comparison with other approaches17:10
Half Sleeves17:35
Complex Background18:20
Unsupervised Creation of Sign Models19:10
Same Sign in Different Contexts19:17
Video19:22
Gestaltic Representation21:26