The authors introduced a new large-scale dataset,the ASL Lexicon Video Dataset [11, 18, 22, 23 and 24],containing video sequences of a large number of distinctsign classes of ASL. This dataset is publicly available nowand expanding rapidly. The authors believe that thisdataset will be an important resource for researchers insign language recognition and human activity analysis, byproviding a large amount of data that can be used fortraining and testing, and by providing a public benchmarkdataset on which different methods can be evaluated. Thedataset is currently a part of a computer vision system thatallows users to look up the meaning of a signautomatically. The authors suggested that the dataset canbe used for testing a wide variety of computer vision,machine learning, and database indexing algorithms. Thesigner performs the sign in front of a camera (or, possibly,in a multi-camera set up), and the computer retrieves anddisplays the most similar signs in the lexicon dataset inthis system.
đang được dịch, vui lòng đợi..
