Teaching a model to watch breaking
Read the full storyA breaking battle moves faster than anyone can type.
Producers spent hours tagging what happens in every clip. I proposed and now run Red Bull’s first video annotation pipeline: a multimodal model labels each move, a person checks the ones it’s unsure of, and the labels reach the live broadcast in real time through an API. It’s 87% accurate across 5,100+ labelled clips, starting with breaking and built to take on other sports.
For product I took it from a proposal to a pilot: the spec, the vendors, the evaluation and the funding.
For innovation Red Bull’s first annotation pipeline, built so other sports can follow.
For teaching A person correcting a model is the clearest way I know to show how AI learns. Try it below.
- Compute and engineering from Google, Microsoft, IBM and NVIDIA
- Our paper, Breakdance Video Classification in the Age of Generative AI, is under review
Try it: be the person in the loop
A minute of a breaking battle, cut into moves. Press play and watch the model label each move as it happens. Then decide the ones it isn’t sure of.
- 0:00Toprock
- 0:07Toprock
- 0:13Footwork
- 0:21
Freeze?, Check
- 0:26Power move
- 0:34
Footwork?, Check
- 0:43Power move
- 0:51
Freeze?, Check
Illustration. An invented round drawn as a skeleton: not Red Bull footage or model output.












