Video Super Resolution NIM

NVIDIA Video Super Resolution (VSR) NIM is an AI inference microservice for upscaling video to higher resolutions. It supports two runtime modes with separate deployment artifacts:

• ST 2110 video workflows: SMPTE ST 2110-20 uncompressed video workflows on NVIDIA Holoscan for Media.
• Compressed video workflows: gRPC, RTP/UDP, and Cloud File workflows for encoded video, controlled through the VSR gRPC API.

Use ST 2110 video workflows when deploying VSR into a Holoscan for Media pipeline for live SMPTE ST 2110-20 uncompressed video. Use compressed video workflows when running the NIM as a container and sending encoded-video requests with gRPC, RTP/UDP, or Cloud File inputs and outputs.

For details about the model used inside this NIM, see the underlying NVIDIA Video Super Resolution model card. The model card covers the video post-processing model used for artifact removal, resolution upscaling, and visual quality enhancement:

NVIDIA Video Super Resolution model card

For deployment and API details, see the Video Super Resolution NIM documentation.