| dc.description.abstract | The internet enables almost anyone to locate content on almost any topic. This ability, however, is not easily available for those who sign. In order to provide resources to those whose primary language is sign language, a digital library, called SLaDL, has been created. In order to ensure maximum efficiency of the video-processor that detects sign language, it is important to check that the program works on all video resolutions. Picture-in-picture videos pose a challenge, as they contain fewer pixels and possess different characteristics than standard webcam sign language videos. However, these videos are very important to test as they are less likely to be retrieved otherwise through tags or other metadata. This project aims to detect and identify sign language in picture-in-picture videos through polar motion profiles, working to expand the corpus of videos on which the processor is successful. | en |
