How Oceaneering Runs Wowza’s Video Intelligence Framework on Live Subsea Video

Oceaneering’s remotely operated vehicles inspect pipelines, anodes, and subsea infrastructure thousands of feet below the ocean surface, producing up to 8 hours of video per dive. They needed a way to monitor that much video without the manual labor. Already streaming live with Wowza Streaming Engine, the team is now building Wowza’s Video Intelligence Framework (VIF) into their pipeline to flag anomalies during live dives and tag moments of interest in recorded footage.

In this webinar, Will Stucky of Oceaneering and Mike Vitale, SVP of Product at Wowza, will cover how that workflow comes together and what it takes to run VIF on your own streams. VIF is now available for Streaming Engine, runs AI inference against the streams you already operate, and outputs detections as timed ID3 metadata, burned-in overlays, JSONL logs, and webhooks.

What you’ll get from this session:

• How Oceaneering routes live subsea streams into detection workflows, all inside their own cloud
• How VIF’s architecture separates streaming from inference, so a saturated GPU never blocks video delivery
• What it takes to deploy VIF in your own setup, from running the included RF-DETR object detection model to loading your own trained weights