ABSTRACT
Many people with some form of hearing loss consider lipreading as their primary mode of day-to-day communication. However, finding resources to learn or improve one's lipreading skills can be challenging. This is further exacerbated in the COVID19 pandemic due to restrictions on direct interactions with peers and speech therapists. Today, online MOOCs platforms like Coursera and Udemy have become the most effective form of training for many types of skill development. However, online lipreading resources are scarce as creating such resources is an extensive process needing months of manual effort to record hired ac-tors. Because of the manual pipeline, such platforms are also limited in vocabulary, supported languages, accents, and speakers and have a high usage cost. In this work, we investigate the possibility of replacing real human talking videos with synthetically generated videos. Synthetic data can easily incorporate larger vocabularies, variations in accent, and even local languages and many speakers. We propose an end-to-end automated pipeline to develop such a platform using state-of-the-art talking head video generator networks, text-to-speech models, and computer vision techniques. We then perform an extensive human evaluation using carefully thought out lipreading exercises to validate the quality of our designed platform against the existing lipreading platforms. Our studies concretely point toward the potential of our approach in developing a large-scale lipreading MOOC platform that can impact millions of people with hearing loss. © 2023 IEEE.
ABSTRACT
This paper presents a new dataset for Kazakh-Russian Sign Language (KRSL) created for the purposes of Sign Language Processing. In 2020, Kazakhstan's schools were quickly switched to online mode due to COVID-19 pandemic. Every working day, the El-arna TV channel was broadcasting video lessons for grades from 1 to 11 with sign language translation. This opportunity allowed us to record a corpus with a large vocabulary and spontaneous SL interpretation. To this end, this corpus contains video recordings of Kazakhstan's online school translated to Kazakh-Russian sign language by 7 interpreters. At the moment we collected and cleaned 890 hours of video material. A custom annotation tool was created to make the process of data annotation simple and easy-to-use by Deaf community. To date, around 325 hours of videos have been annotated with glosses and 4,009 lessons out of 4,547 were transcribed with automatic speech-to-text software. KRSL-OnlineSchool dataset will be made publicly available at https://krslproject.github.io/online-school/. © European Language Resources Association (ELRA), licensed under CC-BY-NC 4.0.