Hi Blue Robotics Community,
I’ve spent the last few months developing a Multi-Layer Auto-Labelling Engine specifically for subsea and maritime imagery. We all know the biggest bottleneck in ROV missions isn’t the dive—it’s the 100+ hours of video you have to manually label afterward.
Why this matters:
-
Speed: Processes a 24-hour dataset in minutes.
-
Zero-Human: No more manual bounding boxes or segmenting.
-
Research Ready: Export directly to YOLO, JSON, or CSV for your models.
See it in Action:
-
Full Sample Gallery: View more processed videos here
-
Try it Yourself: You can test the engine with your own images at demolabelling-production.up.railway.app
I’m looking for a few teams currently working with BlueROV2/BlueBoat who are tired of manual data entry. I will run a sample of your footage through the engine for free to show you the results on your specific use case.
Drop a comment below or DM me if you’d like to see what your data looks like when it’s fully automated!
Best,
Suryansh Singh
Founder, Demo Labelling