NVIDIA Expands AI Tools for Sports Tracking and Broadcast Replay
NVIDIA’s IBC showcase adds single-camera athlete tracking, AI slow-motion replay and HDR conversion, with development through industry partners.
The IBC showcase pairs single-camera motion analysis with AI-generated replay frames and video enhancement.
NVIDIA expanded its AI for Media portfolio at September’s IBC conference in Amsterdam, introducing tools for athlete tracking, slow-motion replay and video enhancement that developers can integrate into sports products.
Ministry of Sport reported October 6 that NVIDIA is pursuing collaborative development rather than selling finished sports applications directly. It named Hawk-Eye, Stats Perform, Peripheral Labs and Skyrim.ai as partners, without specifying individual product launch dates or league deployments.
3D Body Pose estimates human joint positions and angles from a single camera without markers. NVIDIA identifies performance analysis, officiating and player safety among potential sports uses. Its model documentation lists a September 30 release and supports both real-time and offline processing.
Video Frame Generation creates intermediate frames to smooth motion and enable slow-motion playback. NVIDIA says Ross Video is integrating it into Rio Replay, supporting six-times slow motion, with development underway toward eight-times interpolation.
TrueHDR converts standard-dynamic-range video into high-dynamic-range output in real time. It can be combined with frame generation and NVIDIA’s video-upscaling technology in one processing pipeline.
The company also offers Sports Intelligence Playbooks, frameworks for developers to adapt open models using sports footage and annotations. NVIDIA’s IBC announcement identifies Machina Sports as an integrator of those playbooks for rights holders.
The tools span movement data, replay production and sports-specific model development. The announcements do not establish a deployment schedule for the four sports partners named by Ministry of Sport.



