Tuesday, 15 September 2026
Anime, manga, and games, with a take · A Yukimedia publication
← all stories industry 1 sources · 1h ago ·

NVIDIA Expands AI for Media with Frame Generation and Lip Sync Tools

NVIDIA is building out a single toolset for broadcasters that covers capture, motion, upscaling, and audio, which points to production pipelines where real-time translation and dubbing run alongside the video processing itself.

Reporting from 1 source: GIGAZINE.

NVIDIA Expands AI for Media with Frame Generation and Lip Sync Tools

NVIDIA added several AI tools to its AI for Media collection for the broadcast industry in September 2026. The new additions include 3D Body Pose, which converts joint positions into 2D and 3D motion data, Video Frame Generation for raising frame rates two to four times, Video Super Resolution for upscaling, TrueHDR for real-time SDR to HDR conversion, LipSync for matching mouth movements to audio, Active Speaker Detection, and Studio Voice for reducing noise and reverberation.

NVIDIA 3D Body Pose estimates joint positions and angles and turns them into structured 2D and 3D data without markers on the subject, which the source suggests for tracking athletes during replays and for animation production. Video Frame Generation inserts generated frames between existing ones to raise frame rates by two to four times, and work continues toward 8x slow motion. Video Super Resolution upscales while reducing noise, blur, and compression artifacts, with a real-time mode and a quality mode. TrueHDR converts SDR to HDR in real time at up to about 2000 nits.

  • NVIDIA 3D Body Pose: Estimates joint positions and angles as structured 2D and 3D data without markers
  • Video Frame Generation (VFG): Inserts generated frames to raise frame rates two to four times
  • NVIDIA Video Super Resolution (VSR): Upscales video and reduces noise, blur, and compression artifacts
  • NVIDIA TrueHDR: Converts SDR to HDR in real time at up to about 2000 nits
  • NVIDIA LipSync: Changes mouth movements to match input audio
  • Active Speaker Detection: Identifies multiple speakers in an audio file
  • Studio Voice: Reduces noise and reverberation in recorded data

Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.

NVIDIA AI for Media NVIDIA

More on this

  • industry JGBA To Hold 4th Supporting Member General Meeting On November 12

    JGBA is holding a members-only networking event as its member company count approaches 500, which the association frames as the reason for an evening built around face-to-face contact between companies that rarely talk.

  • industry Comunii Selected for Tokyo Government's APT Women Program

    The selection puts a PTA operations tool inside a municipal program built to grow women-led ventures, giving Comunii access to the management training and partner networks it says it wants for expansion.

  • industry Yoom Raises 700 Million Yen In First External Funding Round

    The round gives Yoom money to push its agent platform into core corporate systems, and it arrives with Canon Marketing Japan's investment arm among the backers, which points to distribution interest from a large office-equipment and IT reseller.

  • industry AXLBIT To Bring Japan-Made Software To Tech Week Singapore

    AXLBIT is selling Japan-made software abroad at a time when the country's digital-related trade deficit reached roughly 6.8 trillion yen in 2025, so the pitch is less about a single trade show booth than about whether a one-stop sales, payment, and support structure can move Japanese vendors into ASEAN without each one building a local subsidiary.

  • industry Musk's X And SpaceXAI Withdraw Antitrust Suit Against Apple

    The withdrawal leaves X's broader claim that Apple rigged App Store rankings untested in court, while the parallel OpenAI case over ChatGPT in Apple Intelligence proceeds.

Sources