Computer science
Natural language processing
We are conducting R&D on a program production support system that analyzes large amounts of text data on social media and in program archives within and outside the studio, and is able to extract and present useful information in program production. We have also built a Japanese-English bilingual corpus from news manuscripts, and develop high-quality Japanese-English machine translation for news.
Image/audio analysis
We are also conducting R&D on technology able to automatically read text appearing in video, and video classification technology able to identify names of people and things and also their behaviors and actions. We are working to improve speech recognition technology to transcribe speech from recorded audio, and are pursuing R&D on technology to automatically generate summary videos based on characteristics of video and speech taken from program video.
