training?

[alicera]

1. RetinaFaceAntiCov
2. coordinateReg

can these model train? Or it only provide test?

[nttstar]

Yes, test only.

[alicera]

the coordinateReg

''' self.detector = insightface.model_zoo.get_model('retinaface_mnet025_v2') #can replace with your own face detector #self.detector = insightface.model_zoo.get_model('retinaface_r50_v1') self.detector.prepare(ctx_id=ctx_id) self.det_size = det_size sym, arg_params, aux_params = mx.model.load_checkpoint(prefix, epoch) all_layers = sym.get_internals() sym = all_layers['fc1_output'] self.image_size = image_size model = mx.mod.Module(symbol=sym, context=ctx, label_names = None) model.bind(for_training=False, data_shapes=[('data', (1, 3, image_size[0], image_size[1]))]) model.set_params(arg_params, aux_params) self.model = model '''

The code say the coordinateReg model replaces the the fc1_output of retinaface_mnet025_v2 ? so coordinateReg is just a classifier ?

[kiashann]

1. RetinaFaceAntiCov
2. coordinateReg

can these model train? Or it only provide test?

Did you find any document or example for training or fine-tuning models for face recognition?