search

yinguobing / head-pose-estimation

Realtime human head pose estimation with ONNXRuntime and OpenCV.
MIT License
1.24k stars 288 forks source link

Training and adding another class ?? #6

Closed Abduoit closed 5 years ago

Abduoit commented 6 years ago

Thanks for sharing your great work

Can I replace the file frozen_inference_graph.pb with another one that I trained my model on. Or can I re-train it for another class besides face.

Thanks ??

yinguobing commented 6 years ago

I think you can do either if it suits your needs. Also be awared that the 3D model should also be modified.

This post could be helpful: http://www.learnopencv.com/head-pose-estimation-using-opencv-and-dlib/

devanandd8245 commented 6 years ago

I would like to know whether i can train my dataset and load it in the model file frozen_inference_graph.pb. Could you please help me with some steps to be followed for generating the frozen_inference_graph.pb model file.

yinguobing commented 6 years ago

This could be helpful: https://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/tools/freeze_graph.py

devanandd8245 commented 6 years ago

Please could you help me with steps to be followed as i am new to TensorFlow.

Steps to be followed is that what are the files to be created and how to place them in sequence

devanandd8245 commented 6 years ago

I need the procedure to be followed for the training. Kindly guide me with the steps to be followed for the training module.

yinguobing commented 6 years ago

For exporting model, this could be a good start point: https://yinguobing.com/facial-landmark-localization-by-deep-learning-save-model-application/

devanandd8245 commented 6 years ago

Could you send me the English version page as i am not able to understand the Chinese language

yinguobing commented 6 years ago

Sorry there is only Chinese version available. I'm afraid you have to try Google translate or, ask a friend who happens to understand Chinese for help.

yinguobing commented 5 years ago

Closed. For model training please refer to the other repository: cnn-facial-landmark.