Case study
Thermal Video
3.5 Million Objects Tracked, at 99% Quality
3.5M+
objects tracked
290,183
frames annotated
99%
quality
4 days
to a live 45-person team
Client confidential
The Challenge
A confidential client needed people detected and tracked frame by frame across thermal video, with incidents marked accurately.
What We Did
We sourced, trained and launched a 45-person team in 4 days, covering bounding boxes, frame-to-frame tracking and attribute identification.
The Result
3,531,460 objects tracked across 290,183 frames, delivered at 99% quality.