# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.
import os
import h5py
import numpy as np
from keras.models import Model
def loaddata(filename):
file = h5py.File(filename, 'r')
train_x = file['x'][:]
train_y = file['y'][:]
valid_x = file['valid_x'][:]
valid_y = file['valid_y'][:]
test_x = file['test_x'][:]
test_y = file['test_y'][:]
file.close()
return train_x, train_y, valid_x, valid_y, test_x, test_y
def load_NTU(case):
if case == 0:
filename = os.path.join('./data/ntu', 'NTU_CS.h5')
else:
filename = os.path.join('./data/NTU', 'NTU_CV.h5')
train_x, train_y, valid_x, valid_y, test_x, test_y = loaddata(filename)
return train_x, train_y, valid_x, valid_y, test_x, test_y
def get_data(dataset, case):
if dataset == 'NTU':
train_x, train_y, valid_x, valid_y, test_x, test_y = load_NTU(case)
return train_x, train_y, valid_x, valid_y, test_x, test_y
def Downsample(train_x, valid_x, test_x, ratio):
train_x = train_x[:,0::ratio,:]
valid_x = valid_x[:,0::ratio,:]
test_x = test_x[:, 0::ratio, :]
return train_x, valid_x, test_x
def get_cases(dataset):
if dataset == 'NTU':
cases = 2
return cases
def get_activation(model,test_x, test_y, path, VA = 11 , par = 10):
intermediat_layer_model = Model(inputs= model.input, outputs=model.layers[-2].output)
score = intermediat_layer_model.predict(test_x, batch_size = 256)
pred = softmax(score)
pred_label = np.argmax(pred, axis= -1)
label = np.argmax(test_y, axis= -1)
total = ((label-pred_label)==0).sum()
print("Model Accuracy:%.2f" % (float(total) / len(label)*100))
np.savetxt(path, score, fmt = '%f')
return (float(total)/len(label)*100)
def softmax(data):
e = np.exp(data - np.amax(data, axis=-1, keepdims=True))
s = np.sum(e, axis=-1, keepdims=True)
return e / s
