def vgg_net(weights, image): layers = ( 'conv1_1', 'relu1_1', 'conv1_2', 'relu1_2', 'pool1', 'conv2_1', 'relu2_1', 'conv2_2', 'relu2_2', 'pool2', 'conv3_1', 'relu3_1', 'conv3_2', 'relu3_2', 'conv3_3', 'relu3_3', 'conv3_4', 'relu3_4', 'pool3', 'conv4_1', 'relu4_1', 'conv4_2', 'relu4_2', 'conv4_3', 'relu4_3', 'conv4_4', 'relu4_4', 'pool4', 'conv5_1', 'relu5_1', 'conv5_2', 'relu5_2', 'conv5_3', 'relu5_3', 'conv5_4', 'relu5_4' ) net = {} current = image for i, name in enumerate(layers): kind = name[:4] if kind == 'conv': kernels, bias = weights[i][0][0][0][0] # matconvnet: weights are [width, height, in_channels, out_channels] # tensorflow: weights are [height, width, in_channels, out_channels] if name == 'conv1_1': np.random.seed(3796) append_channels= np.random.normal(loc=0,scale=0.02,size=(3,3,2,64)) # print(append_channels) kernels = np.concatenate((kernels, append_channels), axis=2) kernels = utils.get_variable(kernels, name=name + "_w") else: kernels = utils.get_variable(kernels, name=name + "_w") bias = utils.get_variable(bias.reshape(-1), name=name + "_b") current = utils.conv2d_basic(current, kernels, bias) elif kind == 'relu': current = tf.nn.relu(current, name=name) if FLAGS.debug: utils.add_activation_summary(current) elif kind == 'pool': current = utils.avg_pool_2x2(current) net[name] = current return net
def inference(image, keep_prob): """ Semantic segmentation network definition :param image: input image. Should have values in range 0-255 :param keep_prob: :return: """ print("setting up vgg initialized conv layers ...") model_data = utils.get_model_data(FLAGS.model_dir) mean = model_data['normalization'][0][0][0] mean_pixel = np.mean(mean, axis=(0, 1)) mean_pixel = np.append(mean_pixel, [ 30.6986130799, 284.97018, 106.314329243, 124.171918054, 109.260369903, 182.615729022, 75.1762766769, 84.3529895303, 100.699252985, 66.8837693324, 98.6030061849, 133.955897217 ]) weights = np.squeeze(model_data['layers']) processed_image = utils.process_image(image, mean_pixel) with tf.variable_scope("inference"): image_net = vgg_net(weights, processed_image) conv_final_layer = image_net["conv5_3"] pool5 = utils.max_pool_2x2(conv_final_layer) W6 = utils.weight_variable([7, 7, 512, 4096], name="W6") b6 = utils.bias_variable([4096], name="b6") conv6 = utils.conv2d_basic(pool5, W6, b6) relu6 = tf.nn.relu(conv6, name="relu6") if FLAGS.debug: utils.add_activation_summary(relu6) relu_dropout6 = tf.nn.dropout(relu6, keep_prob=keep_prob) W7 = utils.weight_variable([1, 1, 4096, 4096], name="W7") b7 = utils.bias_variable([4096], name="b7") conv7 = utils.conv2d_basic(relu_dropout6, W7, b7) relu7 = tf.nn.relu(conv7, name="relu7") if FLAGS.debug: utils.add_activation_summary(relu7) relu_dropout7 = tf.nn.dropout(relu7, keep_prob=keep_prob) W8 = utils.weight_variable([1, 1, 4096, NUM_OF_CLASSESS], name="W8") b8 = utils.bias_variable([NUM_OF_CLASSESS], name="b8") conv8 = utils.conv2d_basic(relu_dropout7, W8, b8) # annotation_pred1 = tf.argmax(conv8, dimension=3, name="prediction1") # now to upscale to actual image size deconv_shape1 = image_net["pool4"].get_shape() W_t1 = utils.weight_variable( [4, 4, deconv_shape1[3].value, NUM_OF_CLASSESS], name="W_t1") b_t1 = utils.bias_variable([deconv_shape1[3].value], name="b_t1") conv_t1 = utils.conv2d_transpose_strided(conv8, W_t1, b_t1, output_shape=tf.shape( image_net["pool4"])) fuse_1 = tf.add(conv_t1, image_net["pool4"], name="fuse_1") deconv_shape2 = image_net["pool3"].get_shape() W_t2 = utils.weight_variable( [4, 4, deconv_shape2[3].value, deconv_shape1[3].value], name="W_t2") b_t2 = utils.bias_variable([deconv_shape2[3].value], name="b_t2") conv_t2 = utils.conv2d_transpose_strided(fuse_1, W_t2, b_t2, output_shape=tf.shape( image_net["pool3"])) fuse_2 = tf.add(conv_t2, image_net["pool3"], name="fuse_2") shape = tf.shape(image) deconv_shape3 = tf.stack( [shape[0], shape[1], shape[2], NUM_OF_CLASSESS]) W_t3 = utils.weight_variable( [16, 16, NUM_OF_CLASSESS, deconv_shape2[3].value], name="W_t3") b_t3 = utils.bias_variable([NUM_OF_CLASSESS], name="b_t3") conv_t3 = utils.conv2d_transpose_strided(fuse_2, W_t3, b_t3, output_shape=deconv_shape3, stride=8) annotation_pred = tf.argmax(conv_t3, axis=3, name="prediction") return tf.expand_dims(annotation_pred, dim=3), conv_t3
def inference(image, keep_prob): """ Semantic segmentation network definition :param image: input image. Should have values in range 0-255 :param keep_prob: :return: """ print("setting up vgg pretrained conv layers ...") model_data = utils.get_model_data(FLAGS.model_dir) mean_pixel_init = model_data['normalization'][0][0][0] # mean_pixel_init = np.mean(mean, axis=(0, 1)) mean_pixel = np.zeros((IMAGE_SIZE, IMAGE_SIZE, 6)) mean_pixel[:, :, 0] = mean_pixel_init[:, :, 0] mean_pixel[:, :, 1] = mean_pixel_init[:, :, 1] mean_pixel[:, :, 2] = mean_pixel_init[:, :, 2] mean_pixel[:, :, 3] = np.ones((IMAGE_SIZE, IMAGE_SIZE)) * 97.639895122076 mean_pixel[:, :, 4] = np.ones( (IMAGE_SIZE, IMAGE_SIZE)) * 45.548982715963994 mean_pixel[:, :, 5] = np.ones((IMAGE_SIZE, IMAGE_SIZE)) * 37.69138 weights = np.squeeze(model_data['layers']) processed_image = utils.process_image(image, mean_pixel) with tf.variable_scope("inference"): image_net = vgg_net(weights, processed_image) conv_final_layer = image_net["conv5_3"] pool5 = utils.max_pool_2x2(conv_final_layer) W6 = utils.weight_variable([7, 7, 512, 4096], name="W6") b6 = utils.bias_variable([4096], name="b6") conv6 = utils.conv2d_basic(pool5, W6, b6) relu6 = tf.nn.relu(conv6, name="relu6") if FLAGS.debug: utils.add_activation_summary(relu6) relu_dropout6 = tf.nn.dropout(relu6, keep_prob=keep_prob) W7 = utils.weight_variable([1, 1, 4096, 4096], name="W7") b7 = utils.bias_variable([4096], name="b7") conv7 = utils.conv2d_basic(relu_dropout6, W7, b7) relu7 = tf.nn.relu(conv7, name="relu7") if FLAGS.debug: utils.add_activation_summary(relu7) relu_dropout7 = tf.nn.dropout(relu7, keep_prob=keep_prob) W8 = utils.weight_variable([1, 1, 4096, NUM_OF_CLASSES], name="W8") b8 = utils.bias_variable([NUM_OF_CLASSES], name="b8") conv8 = utils.conv2d_basic(relu_dropout7, W8, b8) annotation_pred1 = tf.argmax(conv8, axis=3, name="prediction1") # now to upscale to actual image size deconv_shape1 = image_net["pool4"].get_shape() W_t1 = utils.weight_variable( [4, 4, deconv_shape1[3].value, NUM_OF_CLASSES], name="W_t1") b_t1 = utils.bias_variable([deconv_shape1[3].value], name="b_t1") conv_t1 = utils.conv2d_transpose_strided(conv8, W_t1, b_t1, output_shape=tf.shape( image_net["pool4"])) fuse_1 = tf.add(conv_t1, image_net["pool4"], name="fuse_1") deconv_shape2 = image_net["pool3"].get_shape() W_t2 = utils.weight_variable( [4, 4, deconv_shape2[3].value, deconv_shape1[3].value], name="W_t2") b_t2 = utils.bias_variable([deconv_shape2[3].value], name="b_t2") conv_t2 = utils.conv2d_transpose_strided(fuse_1, W_t2, b_t2, output_shape=tf.shape( image_net["pool3"])) fuse_2 = tf.add(conv_t2, image_net["pool3"], name="fuse_2") shape = tf.shape(image) deconv_shape3 = tf.stack( [shape[0], shape[1], shape[2], NUM_OF_CLASSES]) W_t3 = utils.weight_variable( [16, 16, NUM_OF_CLASSES, deconv_shape2[3].value], name="W_t3") b_t3 = utils.bias_variable([NUM_OF_CLASSES], name="b_t3") conv_t3 = utils.conv2d_transpose_strided(fuse_2, W_t3, b_t3, output_shape=deconv_shape3, stride=8) annotation_pred = tf.argmax(conv_t3, axis=3, name="prediction") return tf.expand_dims(annotation_pred, dim=3), conv_t3
def inference(image, keep_prob): """ Semantic segmentation network definition :param image: input image. Should have values in range 0-255 :param keep_prob: :return: """ print("setting up vgg pretrained conv layers ...") model_data = utils.get_model_data(FLAGS.model_dir) mean = model_data['normalization'][0][0][0] mean_pixel = np.mean(mean, axis=(0, 1)) mean_pixel = np.append(mean_pixel, [30.69861307993539, 284.9702]) # mean_pixel = np.array([120.8952399852595, 81.93008162338278, 81.28988761879855, 30.69861307993539, 284.9702]) weights = np.squeeze(model_data['layers']) processed_image = utils.process_image(image, mean_pixel) with tf.variable_scope("inference"): net = vgg_net(weights, processed_image) conv_final_layer = net["conv5_3"] pool5 = utils.max_pool_2x2(conv_final_layer) W6 = utils.weight_variable([7, 7, 512, 4096], name="W6") b6 = utils.bias_variable([4096], name="b6") conv6 = utils.conv2d_basic(pool5, W6, b6) relu6 = tf.nn.relu(conv6, name="relu6") if FLAGS.debug: utils.add_activation_summary(relu6) relu_dropout6 = tf.nn.dropout(relu6, keep_prob=keep_prob) W7 = utils.weight_variable([1, 1, 4096, 4096], name="W7") b7 = utils.bias_variable([4096], name="b7") conv7 = utils.conv2d_basic(relu_dropout6, W7, b7) relu7 = tf.nn.relu(conv7, name="relu7") if FLAGS.debug: utils.add_activation_summary(relu7) relu_dropout7 = tf.nn.dropout(relu7, keep_prob=keep_prob) W8 = utils.weight_variable([1, 1, 4096, NUM_OF_CLASSES], name="W8") b8 = utils.bias_variable([NUM_OF_CLASSES], name="b8") conv8 = utils.conv2d_basic(relu_dropout7, W8, b8) annotation_pred1 = tf.argmax(conv8, axis=3, name="prediction1") # now to upscale to actual image size deconv_shape1 = net["pool4"].get_shape() W_t1 = utils.weight_variable([4, 4, deconv_shape1[3].value, NUM_OF_CLASSES], name="W_t1") b_t1 = utils.bias_variable([deconv_shape1[3].value], name="b_t1") conv_t1 = utils.conv2d_transpose_strided(conv8, W_t1, b_t1, output_shape=tf.shape(net["pool4"])) fuse_1 = tf.add(conv_t1, net["pool4"], name="fuse_1") deconv_shape2 = net["pool3"].get_shape() W_t2 = utils.weight_variable([4, 4, deconv_shape2[3].value, deconv_shape1[3].value], name="W_t2") b_t2 = utils.bias_variable([deconv_shape2[3].value], name="b_t2") conv_t2 = utils.conv2d_transpose_strided(fuse_1, W_t2, b_t2, output_shape=tf.shape(net["pool3"])) fuse_2 = tf.add(conv_t2, net["pool3"], name="fuse_2") deconv_shape3 = net["pool2"].get_shape() W_t3 = utils.weight_variable([4, 4, deconv_shape3[3].value, deconv_shape2[3].value], name="W_t3") b_t3 = utils.bias_variable([deconv_shape3[3].value], name="b_t3") conv_t3 = utils.conv2d_transpose_strided(fuse_2, W_t3, b_t3, output_shape=tf.shape(net["pool2"])) fuse_3 = tf.add(conv_t3, net["pool2"], name="fuse_3") shape = tf.shape(image) deconv_shape4 = tf.stack([shape[0], shape[1], shape[2], NUM_OF_CLASSES]) W_t4 = utils.weight_variable([8, 8, NUM_OF_CLASSES, deconv_shape3[3].value], name="W_t4") b_t4 = utils.bias_variable([NUM_OF_CLASSES], name="b_t4") conv_t4 = utils.conv2d_transpose_strided(fuse_3, W_t4, b_t4, output_shape=deconv_shape4, stride=4) annotation_pred = tf.argmax(conv_t4, axis=3, name="prediction") return tf.expand_dims(annotation_pred, dim=3), conv_t4