search

DefTruth / lite.ai.toolkit

🛠 A lite C++ toolkit of awesome AI models, support ONNXRuntime, MNN. Contains YOLOv5, YOLOv6, YOLOX, YOLOv8, FaceDet, HeadSeg, HeadPose, Matting etc. Engine: ONNXRuntime, MNN.
https://github.com/DefTruth/lite.ai.toolkit
GNU General Public License v3.0
3.53k stars 672 forks source link

PipNet access #370

Closed kiralygomba closed 1 year ago

kiralygomba commented 1 year ago

Hello!

Amazing work, thank you! I am using your RetinaFace onnx model and I have already tried a number of facial landmark detectors (mine own and some from you), but I really need the PipNet onnx files. Unfortunately from Europe I cannot download in any way from Baidu. I even tried the free downloaders and registering from a Jordan phonenumber, but no success. Could you upload the PipNet onnx files to Google Drive or add them to docker? Thank you!

DefTruth commented 1 year ago

also available at https://github.com/DefTruth/torchlm/releases/tag/torchlm-0.1.6-alpha

kiralygomba commented 1 year ago

Thanks. For future reference: what I did was use torchIm's method to convert and create the desired .onnx file (watch out for the parameters to match - backbone, number of landmark...etc.) and I could use that directly in lite.ai.toolkit.