Workflow

A standard workflow consists of the following steps, each of them can be triggered by api and the intermediate results can be obtained:

Input Acquisition

Videos can be ingested into the platform from various devices and sources, including mobile phones, cameras, and live streams. This versatility ensures that users can process content from any medium.

Detecting Classes

The platform supports multiple classes for detection, see Obection Detection - Model Zoo

Classifying & Tracking

Once the video is ingested, the platform's tracker and re-identification models classify and track the subjects throughout the video.

Identify Actions

The action detection module identifies key actions and events within the video, providing crucial information for further processing.

Metadata Generation

Detailed metadata is generated, including video intelligence, camera tracking for following objects, highlighting key moments, applying effects, and more. This metadata is essential for organizing and enhancing the video content.

Video Delivery

The final step involves delivering the processed video content. This includes clip grouping, clip output, and output delivery, ensuring that the final product is ready for publishing or further editing.

Compliance and Data Protection

The play Video Intelligence Platform prioritizes compliance and data protection, ensuring that all processed data is handled securely and in accordance with relevant regulations.