def testBuildNonExistingLayerLargeModel(self): """Tests that the model is built correctly without unnecessary layers.""" inputs = tf.random_uniform((5, 331, 331, 3)) tf.train.create_global_step() with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): pnasnet.build_pnasnet_large(inputs, 1000) vars_names = [x.op.name for x in tf.trainable_variables()] self.assertIn('cell_stem_0/1x1/weights', vars_names) self.assertNotIn('cell_stem_1/comb_iter_0/right/1x1/weights', vars_names)
def testBuildPreLogitsLargeModel(self): batch_size = 5 height, width = 331, 331 num_classes = None inputs = tf.random_uniform((batch_size, height, width, 3)) tf.train.create_global_step() with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): net, end_points = pnasnet.build_pnasnet_large(inputs, num_classes) self.assertFalse('AuxLogits' in end_points) self.assertFalse('Predictions' in end_points) self.assertTrue(net.op.name.startswith('final_layer/Mean')) self.assertListEqual(net.get_shape().as_list(), [batch_size, 4320])
def testOverrideHParamsLargeModel(self): batch_size = 5 height, width = 331, 331 num_classes = 1000 inputs = tf.random_uniform((batch_size, height, width, 3)) tf.train.create_global_step() config = pnasnet.large_imagenet_config() config.set_hparam('data_format', 'NCHW') with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): _, end_points = pnasnet.build_pnasnet_large( inputs, num_classes, config=config) self.assertListEqual( end_points['Stem'].shape.as_list(), [batch_size, 540, 42, 42])
def testNoAuxHeadLargeModel(self): batch_size = 5 height, width = 331, 331 num_classes = 1000 for use_aux_head in (True, False): tf.reset_default_graph() inputs = tf.random_uniform((batch_size, height, width, 3)) tf.train.create_global_step() config = pnasnet.large_imagenet_config() config.set_hparam('use_aux_head', int(use_aux_head)) with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): _, end_points = pnasnet.build_pnasnet_large(inputs, num_classes, config=config) self.assertEqual('AuxLogits' in end_points, use_aux_head)
def pnasnet_large_arg_scope_for_detection(is_batch_norm_training=False): """Defines the default arg scope for the PNASNet Large for object detection. This provides a small edit to switch batch norm training on and off. Args: is_batch_norm_training: Boolean indicating whether to train with batch norm. Returns: An `arg_scope` to use for the PNASNet Large Model. """ imagenet_scope = pnasnet.pnasnet_large_arg_scope() with arg_scope(imagenet_scope): with arg_scope([slim.batch_norm], is_training=is_batch_norm_training) as sc: return sc
def testBuildLogitsLargeModel(self): batch_size = 5 height, width = 331, 331 num_classes = 1000 inputs = tf.random_uniform((batch_size, height, width, 3)) tf.train.create_global_step() with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): logits, end_points = pnasnet.build_pnasnet_large(inputs, num_classes) auxlogits = end_points['AuxLogits'] predictions = end_points['Predictions'] self.assertListEqual(auxlogits.get_shape().as_list(), [batch_size, num_classes]) self.assertListEqual(logits.get_shape().as_list(), [batch_size, num_classes]) self.assertListEqual(predictions.get_shape().as_list(), [batch_size, num_classes])
def testAllEndPointsShapesLargeModel(self): batch_size = 5 height, width = 331, 331 num_classes = 1000 inputs = tf.random_uniform((batch_size, height, width, 3)) tf.train.create_global_step() with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): _, end_points = pnasnet.build_pnasnet_large(inputs, num_classes) endpoints_shapes = {'Stem': [batch_size, 42, 42, 540], 'Cell_0': [batch_size, 42, 42, 1080], 'Cell_1': [batch_size, 42, 42, 1080], 'Cell_2': [batch_size, 42, 42, 1080], 'Cell_3': [batch_size, 42, 42, 1080], 'Cell_4': [batch_size, 21, 21, 2160], 'Cell_5': [batch_size, 21, 21, 2160], 'Cell_6': [batch_size, 21, 21, 2160], 'Cell_7': [batch_size, 21, 21, 2160], 'Cell_8': [batch_size, 11, 11, 4320], 'Cell_9': [batch_size, 11, 11, 4320], 'Cell_10': [batch_size, 11, 11, 4320], 'Cell_11': [batch_size, 11, 11, 4320], 'global_pool': [batch_size, 4320], # Logits and predictions 'AuxLogits': [batch_size, 1000], 'Predictions': [batch_size, 1000], 'Logits': [batch_size, 1000], } self.assertEqual(len(end_points), 17) self.assertItemsEqual(endpoints_shapes.keys(), end_points.keys()) for endpoint_name in endpoints_shapes: tf.logging.info('Endpoint name: {}'.format(endpoint_name)) expected_shape = endpoints_shapes[endpoint_name] self.assertIn(endpoint_name, end_points) self.assertListEqual(end_points[endpoint_name].get_shape().as_list(), expected_shape)
def _extract_box_classifier_features(self, proposal_feature_maps, scope): """Extracts second stage box classifier features. This function reconstructs the "second half" of the PNASNet network after the part defined in `_extract_proposal_features`. Args: proposal_feature_maps: A 4-D float tensor with shape [batch_size * self.max_num_proposals, crop_height, crop_width, depth] representing the feature map cropped to each proposal. scope: A scope name. Returns: proposal_classifier_features: A 4-D float tensor with shape [batch_size * self.max_num_proposals, height, width, depth] representing box classifier features for each proposal. """ del scope # Number of used stem cells. num_stem_cells = 2 # Note that we always feed into 2 layers of equal depth # where the first N channels corresponds to previous hidden layer # and the second N channels correspond to the final hidden layer. hidden_previous, hidden = tf.split(proposal_feature_maps, 2, axis=3) # Note that what follows is largely a copy of build_pnasnet_large() within # pnasnet.py. We are copying to minimize code pollution in slim. # TODO(shlens,skornblith): Determine the appropriate drop path schedule. # For now the schedule is the default (1.0->0.7 over 250,000 train steps). hparams = pnasnet.large_imagenet_config() if not self._is_training: hparams.set_hparam('drop_path_keep_prob', 1.0) # Calculate the total number of cells in the network total_num_cells = hparams.num_cells + num_stem_cells normal_cell = pnasnet.PNasNetNormalCell( hparams.num_conv_filters, hparams.drop_path_keep_prob, total_num_cells, hparams.total_training_steps) with arg_scope([slim.dropout, nasnet_utils.drop_path], is_training=self._is_training): with arg_scope([slim.batch_norm], is_training=self._train_batch_norm): with arg_scope([slim.avg_pool2d, slim.max_pool2d, slim.conv2d, slim.batch_norm, slim.separable_conv2d, nasnet_utils.factorized_reduction, nasnet_utils.global_avg_pool, nasnet_utils.get_channel_index, nasnet_utils.get_channel_dim], data_format=hparams.data_format): # This corresponds to the cell number just past 'Cell_7' used by # _extract_proposal_features(). start_cell_num = 8 true_cell_num = start_cell_num + num_stem_cells with slim.arg_scope(pnasnet.pnasnet_large_arg_scope()): net = _build_pnasnet_base( hidden_previous, hidden, normal_cell=normal_cell, hparams=hparams, true_cell_num=true_cell_num, start_cell_num=start_cell_num) proposal_classifier_features = net return proposal_classifier_features