Accurate.Video Validate runs in your web browser, enabling remote users to easily playback and verify professional media content (frame accurate video with multiple audio and subtitle tracks), alongside time-based metadata, such as manual markers (for collaboration or review & approval), Auto-QC report information, or Machine Learning data.

Accurate.Video Validate works with the Amazon Rekognition Video Media Analysis service.

Rekognition Video Media Analysis uses Machine Learning to automatically detect shot changes in pre-edited content. This is useful for promo editors, and for technical operators setting adbreak markers, for commercial broadcasting, VOD and OTT distribution.

In addition, Rekognition Video Media Analysis supports end credit detection in content, which helps technical operators determine where to add "coming next" markers, for consumer binge watching of VOD/OTT episodic content.

Rekognition Video Media Analysis also supports automatic detection of black frames and color bars, which can be removed during post-processing (transcoding), or by content localisation/versioning editing teams.

Accurate.Video Validate runs on the Elastic Container Service (ECS). Once deployed, users can easily connect to, and ingest media from S3 storage, to generate frame accurate H.264 media proxies, using Elemental MediaConvert, to playback video, audio and subtitle content in the Accurate Video player.

Media proxies can also be automatically sent to Amazon Rekognition Video for Media Analysis, or by manual triggering, from within the Accurate.Video Validate user interface.

Once the Rekognition process is complete, metadata is returned to Accurate.Video Validate, where it can be reviewed and edited, in the Accurate.Video timeline.

Users can then export Rekognition metadata from Accurate.Video Validate to configurable CSV, XML and JSON formats.