def predict_imagenet(filename): test_x = ImageDataGenerator.get_image(filename) X = tf.placeholder(tf.float32, [None, 227, 227, 3]) keep_prob = tf.placeholder(tf.float32) model = AlexNet(X,keep_prob) y_ = model.create_model() prob = tf.nn.softmax(y_) with tf.Session() as sess: sess.run(tf.global_variables_initializer()) model.initial_weights(sess) y_,output = sess.run([y_,prob],feed_dict={X: [test_x], keep_prob: 1}) ImageDataGenerator.predict(output)
def predict_dog_cat(filename): test_x = ImageDataGenerator.get_image(filename) X = tf.placeholder(tf.float32, [None, 227, 227, 3]) keep_prob = tf.placeholder(tf.float32) model = AlexNet(X, keep_prob, 2) y_ = model.create_model() predict = tf.argmax(y_,1) saver = tf.train.Saver() with tf.Session() as sess: sess.run(tf.global_variables_initializer()) ckpt = tf.train.get_checkpoint_state('./ckpt') if ckpt and tf.train.checkpoint_exists(ckpt.model_checkpoint_path): print("Reading model parameters from %s" % ckpt.model_checkpoint_path) saver.restore(sess, ckpt.model_checkpoint_path) else: print("no model") return pred = sess.run(predict,feed_dict={X: [test_x], keep_prob: 1}) print(cat_dog_class_names[pred[0]])