ankkhedia

ankkhedia

## Loading required packages
library("readr")
library("dplyr")
library("mxnet")
library("abind")
mx.set.seed(1234)


## Preprocessing steps
## generating synthetic data

ankkhedia

def get_feature(model,aligned):
    input_blob = np.expand_dims(aligned, axis=0)
    data = mx.nd.array(input_blob)
    db = mx.io.DataBatch(data=(data,))
    model.forward(db, is_train=False)
   
    # Normalise embedding obtained from forward pass to unit vector
    embedding = model.get_outputs()[0].squeeze()
    embedding /= embedding.norm()
    return embedding

ankkhedia

## Download the images 
img1 = mx.test_utils.download(url='https://s3.amazonaws.com/onnx-model-zoo/arcface/player1.jpg')
img2 = mx.test_utils.download(url='https://s3.amazonaws.com/onnx-model-zoo/arcface/player2.jpg')
## Start preprocessing of images 
# runs preprocessing steps on the images
# prepare the face detector model
det_threshold = [0.6, 0.7, 0.8]
mtcnn_path = os.path.join(os.path.dirname('__file__'), 'mtcnn-model')
detector = MtcnnDetector(model_folder=mtcnn_path, ctx=ctx, num_worker=1, accurate_landmark = True, threshold=det_threshold)
# pass the images through face detector model to get bounding box for faces and align the faces

ankkhedia

def get_model(ctx, model):
    image_size = (112,112)
    # Import ONNX model
    sym, arg_params, aux_params = import_model(model)
    # Define and binds parameters to the network
    model = mx.mod.Module(symbol=sym, context=ctx, label_names = None)
    #Bind method requires name of the input('data' in this case) and shape of the input for the model
    model.bind(data_shapes=[('data', (1, 3, image_size[0], image_size[1]))])
    model.set_params(arg_params, aux_params)
    return model

ankkhedia

import mxnet as mx 
##Download helper scripts
mx.test_utils.download(url='https://raw.githubusercontent.com/onnx/models/master/models/face_recognition/ArcFace/mtcnn_detector.py')
mx.test_utils.download(url='https://raw.githubusercontent.com/onnx/models/master/models/face_recognition/ArcFace/helper.py')

## Download ArcFace model
mx.test_utils.download(url='https://s3.amazonaws.com/onnx-model-zoo/arcface/resnet100/resnet100.onnx')

## Download pretrained MTCNN face detector model
for i in range(4):