NETINT Technologies collaborates with OpenAI to enhance the inclusivity and accessibility of live video content, showcasing the new technology at the National Association of Broadcasters Show.
In an innovative leap forward for the broadcasting and streaming industries, NETINT Technologies has introduced a groundbreaking automated subtitling feature that leverages the capabilities of OpenAI’s Whisper automatic speech recognition technology. This new feature is set to significantly enhance the accessibility and inclusivity of live video content by providing real-time captioning.
The integration of Whisper into NETINT’s Bitstreams Edge media processing application, which operates on the robust NETINT Quadra Video Server Ampere Edition, marks a pivotal advancement in live broadcast technology. The Whisper model, which has been trained on a staggering 680,000 hours of multilingual data, brings a high level of accuracy and efficiency to the automated transcription process.
The NETINT Quadra Video Server is equipped with the 96-core Ampere® Altra® CPU, renowned for its powerful and energy-efficient performance. This technological synergy enables the transcription of live broadcasts and video streams in real time, setting new standards in the field with unmatched computing efficiency and performance.
This technological solution is not just a technical achievement but also a beacon for cost-effectiveness. Video services, constantly under pressure to reduce operational expenditures, can now integrate high-quality subtitling without significant financial burden. The server supports up to 30 simultaneously transcoded live channels, each with multiple packaged profiles in High-Efficiency Video Coding (HEVC), Advanced Video Coding (H.264), and AV1 formats, underscoring its versatility and broad applicability.
Moreover, this integration showcases a promising collaboration between NETINT Technologies and Ampere Computing. Sean Varley, Chief Evangelist at Ampere, emphasized that this collaboration combines NETINT’s video processing acceleration with Ampere’s AI technology and high-performance processing. This fusion enables the delivery of real-time video transcription at an unprecedented scale, thereby enhancing content delivery network (CDN) operators’ ability to manage streams more effectively and inclusively.
Alex Liu, Co-founder and COO of NETINT, highlighted the inclusive potential of this technology. By utilizing OpenAI’s advanced Whisper model along with NETINT’s innovative video processing solutions, the industry can expect a new era of efficient and cost-effective video captioning. This move significantly aids in making live video content more accessible to diverse audiences worldwide, supporting multiple languages and adapting to various audio environments seamlessly.
The technology will be publicly demonstrated at the upcoming National Association of Broadcasters Show in Las Vegas, offering industry stakeholders a firsthand look at its capabilities.
NETINT’s latest innovation is a clear reflection of the company’s commitment to pioneering green computing approaches within the video domain. Leveraging Advanced Silicon Chip (ASIC)-based video processing solutions, NETINT continues to push the boundaries of transcoding performance, density, and energy efficiency.
This development not only offers practical solutions to current broadcasting challenges but also sets the stage for future advancements in live video streaming and broadcasting, making content universally accessible and engaging for global audiences.