I want to detect face with ssd-mobilenetv2

As far as I know, there is no training model for faces. So I want to train the ssd-mobilenetv2 model that only detects faces. Can you tell me how to do it in detail? I’m using a jetson tx2.


It’s recommended to use a desktop GPU for training task.

For ssd-mobilenetv2, you can use TensorFlow object detection API for the retraining: