search

emilianavt / OpenSeeFace

Robust realtime face and facial landmark tracking on CPU with Unity integration
BSD 2-Clause "Simplified" License
1.41k stars 151 forks source link

Detection model without onnxruntime #54

Closed johngebbie closed 1 year ago

johngebbie commented 1 year ago

Hi @emilianavt, I'm just using opencv without onnxruntime and got the landmarking working using the unoptimized models you shared in https://github.com/emilianavt/OpenSeeFace/issues/48, which is great.

I just tried the optimized detection model (mnv3_detection_opt.onnx) but not that surprisingly got the same layer error as with the optimized landmarking models.

It'd be fab if you could share the unoptimized detection model too.

emilianavt commented 1 year ago

Please try this: mnv3_detection.zip

johngebbie commented 1 year ago

It loads successfully. You're a star!