Thomas Trinh

@zengxianyu so for each new face will have to generate the .npy right?

@dbogunowicz can the converted onnx model be converted to Qualcomm DLC model to run on Qualcomm hardware accelrated device to have the benefit in inference speed?

@quic-ssiddego any updates?

Thanks, I'll try later. Closed for now~

@Wwupup Hi, I take a look at the [gt2coco.py](https://raw.githubusercontent.com/ShiqiYu/libfacedetection.train/master/tools/gt2coco.py) but the **segmentation** is empty. Where do I get these **segmentation** and how do I append it to the trainset.json? I...

Sorry, but as I look at the file on this line https://github.com/ShiqiYu/libfacedetection.train/blob/master/tools/gt2coco.py#L27 I see that the `"segmentation" = []` hardcoded for empty

> It seems that you want to get annotations about the facial key points Yes, but I have my custom dataset and I want to convert it to the format...

Hey @VainF, any updates on adding examples for detection yet?

When built successfully, I have two static lib (libclandmark & libflandmark). Should I import both of them?

I tried to build the lib using android ndk. An error apprears: `'clandmark::CDAGMaxSumSolver' must explicitly initialize the base class 'clandmark::CMaxSumSolver' which does not have a default constructor`?