Pose Estimation Pipeline SDK
Generate keypoints for pose estimation with just 2 lines of code!
Full Documentation DemoSpeed
Works super real-time even with multiple persons in a frame on consumer-grade CPUs.
Robust
Dedicated person-detector enables to estimate Poses robustly in dynamic real-life scenarios.
End to End
Expects rgb[Uint8] data in a Numpy array and returns bounding-boxes in absolute coordinates. Made for developers and creators.
Versatile
No versioning constraints, just Python3 and Numpy. Run it anywhere from a VPS to a local system.
Latency
Person-detection part of pipeline runs in constant time, but pose keypoints would have to calculated separately for each person-detected leading to higher latency. See benchmarks.
Architecture | OS | Time*(ms) |
---|---|---|
Intel i5-8300H Cpu @ 2.30GHz | Windows | 26-27 ms |
NVIDIA GTX 1050 | Windows | ~ 19 ms |
* Time measured in python runtime averaging over 100 loops.
* Time measured with HD resolution as input.