Development of high-performance language processing technologies for audiovisual transcription and translation

Abstract

Natural language processing is a field of computer science, artificial intelligence and linguistics that deals with the interaction between computers and human (natural) languages.

The MLLP group’s research focuses on the development of technology for natural language processing in a large number of languages. The main goal is to facilitate multilingual online communication to overcome language barriers in natural language tasks.

The MLLP group enables multilingual online communication by applying complementary natural language processing tasks such as automatic speech recognition, speech synthesis and statistical machine translation. The MLLP group develops specific tailored models that improve performance, and human-machine interaction environments that allow for efficient and cost-effective refinement of the automatic output.

The tools developed by the MLLP group are being applied to the automatic generation of quality transcripts and translations in video chat repositories and, as a result, advanced features in digital content management, such as classification, recommendation, fragmentation, etc., have been derived. The contexts in which such transcriptions or subtitles are desirable and necessary are numerous: online transcription services, dictation, TV subtitling, language learning, and many others.

The MLLP group develops systems for use in real-life natural language processing applications. These systems are distributed publicly under open source licenses. The performance of the open source tools developed by the MLLP group can be optimized by providing consulting services to interested customers.

The members of the MLLP group have acquired extensive experience in the development of applications related to natural language processing through their participation in numerous publicly and privately funded national and international research projects.

Scientific officer

Juan Císcar Alfonso

Stakeholders

Applications

  • Automatic generation of quality transcriptions and translations in video repositories: online transcription services, dictation, subtitling for TV, language learning…

Technical advantages

  • Developed environments enable efficient and cost-effective refinement of automated output

Benefits it provides

  • High quality transcriptions and translations