Open Hmamouche opened 4 years ago
I just had a look and the model is indeed unable to track the face, it is because the face detection on the video fails. You can change the face detector being used to the default OpenCV one or the dlib one in the code, that might work.
e.g. adding the following line:
det_parameters.curr_face_detector = LandmarkDetector::FaceModelParameters::HOG_SVM_DETECTOR
Thank you.
Hi,
I tried to used FeatureExtraction and/or FaceLandmarkVid on a video containing images of a robot. The models don't detect landmarks. I tried all the available models, but without a results. You can find the video in this Onedrive link: https://1drv.ms/v/s!AlweC-qetX4AgZhFtgNQhPj4MdRnqw?e=vqGOdx
Best regards,