project_name
stringlengths
6
104
file_name
stringlengths
4
89
full_name
stringlengths
1
102
func_name
stringlengths
1
85
docstring
stringlengths
13
836
docstring_tokens
listlengths
4
122
code
stringlengths
23
39.7k
code_tokens
stringlengths
29
44.6k
url
int64
3
986k
sek788432/Waymo-2D-Object-Detection
eval_util.py
top_k_by_class
top_k_by_class
Extracts the top k predictions for each video, sorted by class.
[ "Extracts", "the", "top", "k", "predictions", "for", "each", "video,", "sorted", "by", "class." ]
def top_k_by_class(predictions, labels, k=20): if k <= 0: raise ValueError('k must be a positive integer.') k = min(k, predictions.shape[1]) num_classes = predictions.shape[1] prediction_triplets = [] for video_index in range(predictions.shape[0]): prediction_triplets.extend(top_k_tr...
['def', 'top_k_by_class(predictions,', 'labels,', 'k=20):', 'if', 'k', '<=', '0:', 'raise', "ValueError('k", 'must', 'be', 'a', 'positive', "integer.')", 'k', '=', 'min(k,', 'predictions.shape[1])', 'num_classes', '=', 'predictions.shape[1]', 'prediction_triplets', '=', '[]', 'for', 'video_index', 'in', 'range(predicti...
973,429
sek788432/Waymo-2D-Object-Detection
eval_util.py
EvaluationMetrics.accumulate
accumulate
Accumulate the metrics calculated locally for this mini-batch.
[ "Accumulate", "the", "metrics", "calculated", "locally", "for", "this", "mini-batch." ]
def accumulate(self, predictions, labels): (predictions, labels) = self._convert_to_numpy(predictions=predictions[0], groundtruths=labels[0]) batch_size = labels.shape[0] mean_hit_at_one = calculate_hit_at_one(predictions, labels) mean_perr = calculate_precision_at_equal_recall_rate(predictions, labels)...
['def', 'accumulate(self,', 'predictions,', 'labels):', '(predictions,', 'labels)', '=', 'self._convert_to_numpy(predictions=predictions[0],', 'groundtruths=labels[0])', 'batch_size', '=', 'labels.shape[0]', 'mean_hit_at_one', '=', 'calculate_hit_at_one(predictions,', 'labels)', 'mean_perr', '=', 'calculate_precision_a...
973,431
sek788432/Waymo-2D-Object-Detection
eval_util.py
EvaluationMetrics.clear
clear
Clear the evaluation metrics and reset the EvaluationMetrics object.
[ "Clear", "the", "evaluation", "metrics", "and", "reset", "the", "EvaluationMetrics", "object." ]
def clear(self): self.sum_hit_at_one = 0.0 self.sum_perr = 0.0 self.map_calculator.clear() self.global_ap_calculator.clear() self.num_examples = 0
['def', 'clear(self):', 'self.sum_hit_at_one', '=', '0.0', 'self.sum_perr', '=', '0.0', 'self.map_calculator.clear()', 'self.global_ap_calculator.clear()', 'self.num_examples', '=', '0']
973,433
sek788432/Waymo-2D-Object-Detection
yt8m_agg_models.py
LogisticModel.create_model
create_model
Creates a logistic model.
[ "Creates", "a", "logistic", "model." ]
def create_model(self, model_input, vocab_size, l2_penalty=1e-08): output = layers.Dense(vocab_size, activation=tf.nn.sigmoid, kernel_regularizer=regularizers.l2(l2_penalty))(model_input) return {'predictions': output}
['def', 'create_model(self,', 'model_input,', 'vocab_size,', 'l2_penalty=1e-08):', 'output', '=', 'layers.Dense(vocab_size,', 'activation=tf.nn.sigmoid,', 'kernel_regularizer=regularizers.l2(l2_penalty))(model_input)', 'return', "{'predictions':", 'output}']
973,436
sek788432/Waymo-2D-Object-Detection
yt8m_model_test.py
YT8MNetworkTest.test_yt8m_network_creation
test_yt8m_network_creation
Test for creation of a YT8M Model.
[ "Test", "for", "creation", "of", "a", "YT8M", "Model." ]
def test_yt8m_network_creation(self, num_frames, feature_dims): input_specs = tf.keras.layers.InputSpec(shape=[num_frames, feature_dims]) num_classes = 3862 model = yt8m_model.YT8MModel(input_params=yt8m_cfg.YT8MTask.model, num_frames=num_frames, num_classes=num_classes, input_specs=input_specs) inputs ...
['def', 'test_yt8m_network_creation(self,', 'num_frames,', 'feature_dims):', 'input_specs', '=', 'tf.keras.layers.InputSpec(shape=[num_frames,', 'feature_dims])', 'num_classes', '=', '3862', 'model', '=', 'yt8m_model.YT8MModel(input_params=yt8m_cfg.YT8MTask.model,', 'num_frames=num_frames,', 'num_classes=num_classes,',...
973,439
sek788432/Waymo-2D-Object-Detection
yt8m_model_utils.py
SampleRandomSequence
SampleRandomSequence
Samples a random sequence of frames of size num_samples.
[ "Samples", "a", "random", "sequence", "of", "frames", "of", "size", "num_samples." ]
def SampleRandomSequence(model_input, num_frames, num_samples): batch_size = tf.shape(model_input)[0] frame_index_offset = tf.tile(tf.expand_dims(tf.range(num_samples), 0), [batch_size, 1]) max_start_frame_index = tf.maximum(num_frames - num_samples, 0) start_frame_index = tf.cast(tf.multiply(tf.random_...
['def', 'SampleRandomSequence(model_input,', 'num_frames,', 'num_samples):', 'batch_size', '=', 'tf.shape(model_input)[0]', 'frame_index_offset', '=', 'tf.tile(tf.expand_dims(tf.range(num_samples),', '0),', '[batch_size,', '1])', 'max_start_frame_index', '=', 'tf.maximum(num_frames', '-', 'num_samples,', '0)', 'start_f...
973,440
sek788432/Waymo-2D-Object-Detection
yt8m_model_utils.py
SampleRandomFrames
SampleRandomFrames
Samples a random set of frames of size num_samples.
[ "Samples", "a", "random", "set", "of", "frames", "of", "size", "num_samples." ]
def SampleRandomFrames(model_input, num_frames, num_samples): batch_size = tf.shape(model_input)[0] frame_index = tf.cast(tf.multiply(tf.random.uniform([batch_size, num_samples]), tf.tile(tf.cast(num_frames, tf.float32), [1, num_samples])), tf.int32) batch_index = tf.tile(tf.expand_dims(tf.range(batch_size)...
['def', 'SampleRandomFrames(model_input,', 'num_frames,', 'num_samples):', 'batch_size', '=', 'tf.shape(model_input)[0]', 'frame_index', '=', 'tf.cast(tf.multiply(tf.random.uniform([batch_size,', 'num_samples]),', 'tf.tile(tf.cast(num_frames,', 'tf.float32),', '[1,', 'num_samples])),', 'tf.int32)', 'batch_index', '=', ...
973,441
sek788432/Waymo-2D-Object-Detection
yt8m_model_utils.py
FramePooling
FramePooling
Pools over the frames of a video.
[ "Pools", "over", "the", "frames", "of", "a", "video." ]
def FramePooling(frames, method): if method == 'average': reduced = tf.reduce_mean(frames, 1) elif method == 'max': reduced = tf.reduce_max(frames, 1) elif method == 'none': feature_size = frames.shape_as_list()[2] reduced = tf.reshape(frames, [-1, feature_size]) else: ...
['def', 'FramePooling(frames,', 'method):', 'if', 'method', '==', "'average':", 'reduced', '=', 'tf.reduce_mean(frames,', '1)', 'elif', 'method', '==', "'max':", 'reduced', '=', 'tf.reduce_max(frames,', '1)', 'elif', 'method', '==', "'none':", 'feature_size', '=', 'frames.shape_as_list()[2]', 'reduced', '=', 'tf.reshap...
973,442
sek788432/Waymo-2D-Object-Detection
yt8m_task.py
YT8MTask.build_model
build_model
Builds model for YT8M Task.
[ "Builds", "model", "for", "YT8M", "Task." ]
def build_model(self): train_cfg = self.task_config.train_data common_input_shape = [None, sum(train_cfg.feature_sizes)] input_specs = tf.keras.layers.InputSpec(shape=[None] + common_input_shape) logging.info('Build model input %r', common_input_shape) model_config = self.task_config.model model...
['def', 'build_model(self):', 'train_cfg', '=', 'self.task_config.train_data', 'common_input_shape', '=', '[None,', 'sum(train_cfg.feature_sizes)]', 'input_specs', '=', 'tf.keras.layers.InputSpec(shape=[None]', '+', 'common_input_shape)', "logging.info('Build", 'model', 'input', "%r',", 'common_input_shape)', 'model_co...
973,443
sek788432/Waymo-2D-Object-Detection
detection.py
DetectionModule.serve
serve
Cast image to float and run inference.
[ "Cast", "image", "to", "float", "and", "run", "inference." ]
def serve(self, images: tf.Tensor): model_params = self.params.task.model with tf.device('cpu:0'): images = tf.cast(images, dtype=tf.float32) images_spec = tf.TensorSpec(shape=self._input_image_size + [3], dtype=tf.float32) num_anchors = model_params.anchor.num_scales * len(model_params....
['def', 'serve(self,', 'images:', 'tf.Tensor):', 'model_params', '=', 'self.params.task.model', 'with', "tf.device('cpu:0'):", 'images', '=', 'tf.cast(images,', 'dtype=tf.float32)', 'images_spec', '=', 'tf.TensorSpec(shape=self._input_image_size', '+', '[3],', 'dtype=tf.float32)', 'num_anchors', '=', 'model_params.anch...
973,447
sek788432/Waymo-2D-Object-Detection
export_base.py
ExportModule.get_inference_signatures
get_inference_signatures
Gets defined function signatures.
[ "Gets", "defined", "function", "signatures." ]
def get_inference_signatures(self, function_keys: Dict[Text, Text]): signatures = {} for (key, def_name) in function_keys.items(): if key == 'image_tensor': input_signature = tf.TensorSpec(shape=[self._batch_size] + [None] * len(self._input_image_size) + [self._num_channels], dtype=tf.uint8)...
['def', 'get_inference_signatures(self,', 'function_keys:', 'Dict[Text,', 'Text]):', 'signatures', '=', '{}', 'for', '(key,', 'def_name)', 'in', 'function_keys.items():', 'if', 'key', '==', "'image_tensor':", 'input_signature', '=', 'tf.TensorSpec(shape=[self._batch_size]', '+', '[None]', '*', 'len(self._input_image_si...
973,448
sek788432/Waymo-2D-Object-Detection
export_tfhub.py
export_model_to_tfhub
export_model_to_tfhub
Export an image classification model to TF-Hub.
[ "Export", "an", "image", "classification", "model", "to", "TF-Hub." ]
def export_model_to_tfhub(params, batch_size, input_image_size, skip_logits_layer, checkpoint_path, export_path): input_specs = tf.keras.layers.InputSpec(shape=[batch_size] + input_image_size + [3]) model = factory.build_classification_model(input_specs=input_specs, model_config=params.task.model, l2_regularize...
['def', 'export_model_to_tfhub(params,', 'batch_size,', 'input_image_size,', 'skip_logits_layer,', 'checkpoint_path,', 'export_path):', 'input_specs', '=', 'tf.keras.layers.InputSpec(shape=[batch_size]', '+', 'input_image_size', '+', '[3])', 'model', '=', 'factory.build_classification_model(input_specs=input_specs,', '...
973,450
sek788432/Waymo-2D-Object-Detection
video_classification.py
VideoClassificationTask.build_model
build_model
Builds video classification model.
[ "Builds", "video", "classification", "model." ]
def build_model(self): common_input_shape = self._get_feature_shape() input_specs = tf.keras.layers.InputSpec(shape=[None] + common_input_shape) logging.info('Build model input %r', common_input_shape) l2_weight_decay = self.task_config.losses.l2_weight_decay l2_regularizer = tf.keras.regularizers.l...
['def', 'build_model(self):', 'common_input_shape', '=', 'self._get_feature_shape()', 'input_specs', '=', 'tf.keras.layers.InputSpec(shape=[None]', '+', 'common_input_shape)', "logging.info('Build", 'model', 'input', "%r',", 'common_input_shape)', 'l2_weight_decay', '=', 'self.task_config.losses.l2_weight_decay', 'l2_r...
973,465
sek788432/Waymo-2D-Object-Detection
main.py
run_executor
run_executor
Runs the object detection model on distribution strategy defined by the user.
[ "Runs", "the", "object", "detection", "model", "on", "distribution", "strategy", "defined", "by", "the", "user." ]
def run_executor(params, mode, checkpoint_path=None, train_input_fn=None, eval_input_fn=None, callbacks=None, prebuilt_strategy=None): if params.architecture.use_bfloat16: tf.compat.v2.keras.mixed_precision.set_global_policy('mixed_bfloat16') model_builder = model_factory.model_generator(params) if ...
['def', 'run_executor(params,', 'mode,', 'checkpoint_path=None,', 'train_input_fn=None,', 'eval_input_fn=None,', 'callbacks=None,', 'prebuilt_strategy=None):', 'if', 'params.architecture.use_bfloat16:', "tf.compat.v2.keras.mixed_precision.set_global_policy('mixed_bfloat16')", 'model_builder', '=', 'model_factory.model_...
973,471
sek788432/Waymo-2D-Object-Detection
base_model.py
Model.model_outputs
model_outputs
Build the model outputs.
[ "Build", "the", "model", "outputs." ]
def model_outputs(self, inputs, mode): return self.build_outputs(inputs, mode)
['def', 'model_outputs(self,', 'inputs,', 'mode):', 'return', 'self.build_outputs(inputs,', 'mode)']
973,506
sek788432/Waymo-2D-Object-Detection
learning_rates.py
learning_rate_generator
learning_rate_generator
The learning rate function generator.
[ "The", "learning", "rate", "function", "generator." ]
def learning_rate_generator(total_steps, params): if params.type == 'step': return StepLearningRateWithLinearWarmup(total_steps, params) elif params.type == 'cosine': return CosineLearningRateWithLinearWarmup(total_steps, params) else: raise ValueError('Unsupported learning rate type...
['def', 'learning_rate_generator(total_steps,', 'params):', 'if', 'params.type', '==', "'step':", 'return', 'StepLearningRateWithLinearWarmup(total_steps,', 'params)', 'elif', 'params.type', '==', "'cosine':", 'return', 'CosineLearningRateWithLinearWarmup(total_steps,', 'params)', 'else:', 'raise', "ValueError('Unsuppo...
973,512
sek788432/Waymo-2D-Object-Detection
factory.py
oln_rpn_head_generator
oln_rpn_head_generator
Generator function for OLN-proposal (OLN-RPN) head architecture.
[ "Generator", "function", "for", "OLN-proposal", "(OLN-RPN)", "head", "architecture." ]
def oln_rpn_head_generator(params): head_params = params.rpn_head anchors_per_location = params.anchor.num_scales * len(params.anchor.aspect_ratios) return heads.OlnRpnHead(params.architecture.min_level, params.architecture.max_level, anchors_per_location, head_params.num_convs, head_params.num_filters, hea...
['def', 'oln_rpn_head_generator(params):', 'head_params', '=', 'params.rpn_head', 'anchors_per_location', '=', 'params.anchor.num_scales', '*', 'len(params.anchor.aspect_ratios)', 'return', 'heads.OlnRpnHead(params.architecture.min_level,', 'params.architecture.max_level,', 'anchors_per_location,', 'head_params.num_con...
973,520
sek788432/Waymo-2D-Object-Detection
factory.py
oln_box_score_head_generator
oln_box_score_head_generator
Generator function for Scoring Fast R-CNN head architecture.
[ "Generator", "function", "for", "Scoring", "Fast", "R-CNN", "head", "architecture." ]
def oln_box_score_head_generator(params): head_params = params.frcnn_head return heads.OlnBoxScoreHead(params.architecture.num_classes, head_params.num_convs, head_params.num_filters, head_params.use_separable_conv, head_params.num_fcs, head_params.fc_dims, params.norm_activation.activation, head_params.use_bat...
['def', 'oln_box_score_head_generator(params):', 'head_params', '=', 'params.frcnn_head', 'return', 'heads.OlnBoxScoreHead(params.architecture.num_classes,', 'head_params.num_convs,', 'head_params.num_filters,', 'head_params.use_separable_conv,', 'head_params.num_fcs,', 'head_params.fc_dims,', 'params.norm_activation.a...
973,522
sek788432/Waymo-2D-Object-Detection
factory.py
shapeprior_head_generator
shapeprior_head_generator
Generator function for shape prior head architecture.
[ "Generator", "function", "for", "shape", "prior", "head", "architecture." ]
def shapeprior_head_generator(params): head_params = params.shapemask_head return heads.ShapemaskPriorHead(params.architecture.num_classes, head_params.num_downsample_channels, head_params.mask_crop_size, head_params.use_category_for_mask, head_params.shape_prior_path)
['def', 'shapeprior_head_generator(params):', 'head_params', '=', 'params.shapemask_head', 'return', 'heads.ShapemaskPriorHead(params.architecture.num_classes,', 'head_params.num_downsample_channels,', 'head_params.mask_crop_size,', 'head_params.use_category_for_mask,', 'head_params.shape_prior_path)']
973,525
sek788432/Waymo-2D-Object-Detection
heads.py
FastrcnnHead.call
call
Box and class branches for the Mask-RCNN model.
[ "Box", "and", "class", "branches", "for", "the", "Mask-RCNN", "model." ]
def call(self, roi_features, is_training=None): with tf.name_scope('fast_rcnn_head'): (_, num_rois, height, width, filters) = roi_features.get_shape().as_list() net = tf.reshape(roi_features, [-1, height, width, filters]) for i in range(self._num_convs): net = self._conv_ops[i](n...
['def', 'call(self,', 'roi_features,', 'is_training=None):', 'with', "tf.name_scope('fast_rcnn_head'):", '(_,', 'num_rois,', 'height,', 'width,', 'filters)', '=', 'roi_features.get_shape().as_list()', 'net', '=', 'tf.reshape(roi_features,', '[-1,', 'height,', 'width,', 'filters])', 'for', 'i', 'in', 'range(self._num_co...
973,528
sek788432/Waymo-2D-Object-Detection
heads.py
MaskrcnnHead.call
call
Mask branch for the Mask-RCNN model.
[ "Mask", "branch", "for", "the", "Mask-RCNN", "model." ]
def call(self, roi_features, class_indices, is_training=None): with tf.name_scope('mask_head'): (_, num_rois, height, width, filters) = roi_features.get_shape().as_list() net = tf.reshape(roi_features, [-1, height, width, filters]) for i in range(self._num_convs): net = self._con...
['def', 'call(self,', 'roi_features,', 'class_indices,', 'is_training=None):', 'with', "tf.name_scope('mask_head'):", '(_,', 'num_rois,', 'height,', 'width,', 'filters)', '=', 'roi_features.get_shape().as_list()', 'net', '=', 'tf.reshape(roi_features,', '[-1,', 'height,', 'width,', 'filters])', 'for', 'i', 'in', 'range...
973,529
sek788432/Waymo-2D-Object-Detection
heads.py
ShapemaskFinemaskHead.decoder_net
decoder_net
Fine mask decoder network architecture.
[ "Fine", "mask", "decoder", "network", "architecture." ]
def decoder_net(self, features, is_training=False): (batch_size, num_instances, height, width, num_channels) = features.get_shape().as_list() features = tf.reshape(features, [batch_size * num_instances, height, width, num_channels]) for i in range(self._num_convs): features = self._fine_class_conv[i...
['def', 'decoder_net(self,', 'features,', 'is_training=False):', '(batch_size,', 'num_instances,', 'height,', 'width,', 'num_channels)', '=', 'features.get_shape().as_list()', 'features', '=', 'tf.reshape(features,', '[batch_size', '*', 'num_instances,', 'height,', 'width,', 'num_channels])', 'for', 'i', 'in', 'range(s...
973,533
sek788432/Waymo-2D-Object-Detection
target_ops.py
ROISampler.call
call
Sample and assign RoIs for training.
[ "Sample", "and", "assign", "RoIs", "for", "training." ]
def call(self, rois, gt_boxes, gt_classes): (sampled_rois, sampled_gt_boxes, sampled_gt_classes, sampled_gt_indices) = assign_and_sample_proposals(rois, gt_boxes, gt_classes, num_samples_per_image=self._num_samples_per_image, mix_gt_boxes=self._mix_gt_boxes, fg_fraction=self._fg_fraction, fg_iou_thresh=self._fg_iou...
['def', 'call(self,', 'rois,', 'gt_boxes,', 'gt_classes):', '(sampled_rois,', 'sampled_gt_boxes,', 'sampled_gt_classes,', 'sampled_gt_indices)', '=', 'assign_and_sample_proposals(rois,', 'gt_boxes,', 'gt_classes,', 'num_samples_per_image=self._num_samples_per_image,', 'mix_gt_boxes=self._mix_gt_boxes,', 'fg_fraction=se...
973,554
sek788432/Waymo-2D-Object-Detection
class_utils.py
coco_split_class_ids
coco_split_class_ids
Return the COCO class split ids based on split name and training mode.
[ "Return", "the", "COCO", "class", "split", "ids", "based", "on", "split", "name", "and", "training", "mode." ]
def coco_split_class_ids(split_name): if split_name == 'all': return [] elif split_name == 'voc': return [1, 2, 3, 4, 5, 6, 7, 9, 16, 17, 18, 19, 20, 21, 44, 62, 63, 64, 67, 72] elif split_name == 'nonvoc': return [8, 10, 11, 13, 14, 15, 22, 23, 24, 25, 27, 28, 31, 32, 33, 34, 35, 36...
['def', 'coco_split_class_ids(split_name):', 'if', 'split_name', '==', "'all':", 'return', '[]', 'elif', 'split_name', '==', "'voc':", 'return', '[1,', '2,', '3,', '4,', '5,', '6,', '7,', '9,', '16,', '17,', '18,', '19,', '20,', '21,', '44,', '62,', '63,', '64,', '67,', '72]', 'elif', 'split_name', '==', "'nonvoc':", '...
973,571
sek788432/Waymo-2D-Object-Detection
classifier_trainer.py
get_dtype_map
get_dtype_map
Returns the mapping from dtype string representations to TF dtypes.
[ "Returns", "the", "mapping", "from", "dtype", "string", "representations", "to", "TF", "dtypes." ]
def get_dtype_map() -> Mapping[str, tf.dtypes.DType]: return {'float32': tf.float32, 'bfloat16': tf.bfloat16, 'float16': tf.float16, 'fp32': tf.float32, 'bf16': tf.bfloat16}
['def', 'get_dtype_map()', '->', 'Mapping[str,', 'tf.dtypes.DType]:', 'return', "{'float32':", 'tf.float32,', "'bfloat16':", 'tf.bfloat16,', "'float16':", 'tf.float16,', "'fp32':", 'tf.float32,', "'bf16':", 'tf.bfloat16}']
973,714
sek788432/Waymo-2D-Object-Detection
classifier_trainer.py
get_image_size_from_model
get_image_size_from_model
If the given model has a preferred image size, return it.
[ "If", "the", "given", "model", "has", "a", "preferred", "image", "size,", "return", "it." ]
def get_image_size_from_model(params: base_configs.ExperimentConfig) -> Optional[int]: if params.model_name == 'efficientnet': efficientnet_name = params.model.model_params.model_name if efficientnet_name in efficientnet_model.MODEL_CONFIGS: return efficientnet_model.MODEL_CONFIGS[effici...
['def', 'get_image_size_from_model(params:', 'base_configs.ExperimentConfig)', '->', 'Optional[int]:', 'if', 'params.model_name', '==', "'efficientnet':", 'efficientnet_name', '=', 'params.model.model_params.model_name', 'if', 'efficientnet_name', 'in', 'efficientnet_model.MODEL_CONFIGS:', 'return', 'efficientnet_model...
973,715
sek788432/Waymo-2D-Object-Detection
classifier_trainer.py
run
run
Runs Image Classification model using native Keras APIs.
[ "Runs", "Image", "Classification", "model", "using", "native", "Keras", "APIs." ]
def run(flags_obj: flags.FlagValues, strategy_override: tf.distribute.Strategy=None) -> Mapping[str, Any]: params = _get_params_from_flags(flags_obj) if params.mode == 'train_and_eval': return train_and_eval(params, strategy_override) elif params.mode == 'export_only': export(params) els...
['def', 'run(flags_obj:', 'flags.FlagValues,', 'strategy_override:', 'tf.distribute.Strategy=None)', '->', 'Mapping[str,', 'Any]:', 'params', '=', '_get_params_from_flags(flags_obj)', 'if', 'params.mode', '==', "'train_and_eval':", 'return', 'train_and_eval(params,', 'strategy_override)', 'elif', 'params.mode', '==', "...
973,723
sek788432/Waymo-2D-Object-Detection
classifier_trainer_test.py
get_params_override
get_params_override
Converts params_override dict to string command.
[ "Converts", "params_override", "dict", "to", "string", "command." ]
def get_params_override(params_override: Mapping[str, Any]) -> str: return '--params_override=' + json.dumps(params_override)
['def', 'get_params_override(params_override:', 'Mapping[str,', 'Any])', '->', 'str:', 'return', "'--params_override='", '+', 'json.dumps(params_override)']
973,725
sek788432/Waymo-2D-Object-Detection
dataset_factory.py
DatasetConfig.has_data
has_data
Whether this dataset is has any data associated with it.
[ "Whether", "this", "dataset", "is", "has", "any", "data", "associated", "with", "it." ]
def has_data(self): return self.name or self.data_dir or self.filenames
['def', 'has_data(self):', 'return', 'self.name', 'or', 'self.data_dir', 'or', 'self.filenames']
973,737
sek788432/Waymo-2D-Object-Detection
dataset_factory.py
DatasetBuilder.image_size
image_size
The size of each image (can be inferred from the dataset).
[ "The", "size", "of", "each", "image", "(can", "be", "inferred", "from", "the", "dataset)." ]
def image_size(self) -> int: if self.config.image_size == 'infer': return self.info.features['image'].shape[0] else: return int(self.config.image_size)
['def', 'image_size(self)', '->', 'int:', 'if', 'self.config.image_size', '==', "'infer':", 'return', "self.info.features['image'].shape[0]", 'else:', 'return', 'int(self.config.image_size)']
973,744
sek788432/Waymo-2D-Object-Detection
dataset_factory.py
DatasetBuilder.num_examples
num_examples
The number of examples (can be inferred from the dataset).
[ "The", "number", "of", "examples", "(can", "be", "inferred", "from", "the", "dataset)." ]
def num_examples(self) -> int: if self.config.num_examples == 'infer': return self.info.splits[self.config.split].num_examples else: return int(self.config.num_examples)
['def', 'num_examples(self)', '->', 'int:', 'if', 'self.config.num_examples', '==', "'infer':", 'return', 'self.info.splits[self.config.split].num_examples', 'else:', 'return', 'int(self.config.num_examples)']
973,746
sek788432/Waymo-2D-Object-Detection
mnist_main.py
build_model
build_model
Constructs the ML model used to predict handwritten digits.
[ "Constructs", "the", "ML", "model", "used", "to", "predict", "handwritten", "digits." ]
def build_model(): image = tf.keras.layers.Input(shape=(28, 28, 1)) y = tf.keras.layers.Conv2D(filters=32, kernel_size=5, padding='same', activation='relu')(image) y = tf.keras.layers.MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same')(y) y = tf.keras.layers.Conv2D(filters=32, kernel_size=5, ...
['def', 'build_model():', 'image', '=', 'tf.keras.layers.Input(shape=(28,', '28,', '1))', 'y', '=', 'tf.keras.layers.Conv2D(filters=32,', 'kernel_size=5,', "padding='same',", "activation='relu')(image)", 'y', '=', 'tf.keras.layers.MaxPooling2D(pool_size=(2,', '2),', 'strides=(2,', '2),', "padding='same')(y)", 'y', '=',...
973,759
sek788432/Waymo-2D-Object-Detection
mnist_test.py
KerasMnistTest.test_end_to_end
test_end_to_end
Test Keras MNIST model with `strategy`.
[ "Test", "Keras", "MNIST", "model", "with", "`strategy`." ]
def test_end_to_end(self, distribution): extra_flags = ['-train_epochs', '1', '--data_dir='] dummy_data = (tf.ones(shape=(10, 28, 28, 1), dtype=tf.int32), tf.range(10)) datasets = (tf.data.Dataset.from_tensor_slices(dummy_data), tf.data.Dataset.from_tensor_slices(dummy_data)) run = functools.partial(mni...
['def', 'test_end_to_end(self,', 'distribution):', 'extra_flags', '=', "['-train_epochs',", "'1',", "'--data_dir=']", 'dummy_data', '=', '(tf.ones(shape=(10,', '28,', '28,', '1),', 'dtype=tf.int32),', 'tf.range(10))', 'datasets', '=', '(tf.data.Dataset.from_tensor_slices(dummy_data),', 'tf.data.Dataset.from_tensor_slic...
973,763
sek788432/Waymo-2D-Object-Detection
optimizer_factory_test.py
OptimizerFactoryTest.test_learning_rate_with_decay_and_warmup
test_learning_rate_with_decay_and_warmup
Basic smoke test for syntax.
[ "Basic", "smoke", "test", "for", "syntax." ]
def test_learning_rate_with_decay_and_warmup(self, lr_decay_type): params = base_configs.LearningRateConfig(name=lr_decay_type, initial_lr=0.01, decay_rate=0.01, decay_epochs=1, warmup_epochs=1, scale_by_batch_size=0.01, examples_per_epoch=1, boundaries=[0], multipliers=[0, 1]) batch_size = 1 train_epochs =...
['def', 'test_learning_rate_with_decay_and_warmup(self,', 'lr_decay_type):', 'params', '=', 'base_configs.LearningRateConfig(name=lr_decay_type,', 'initial_lr=0.01,', 'decay_rate=0.01,', 'decay_epochs=1,', 'warmup_epochs=1,', 'scale_by_batch_size=0.01,', 'examples_per_epoch=1,', 'boundaries=[0],', 'multipliers=[0,', '1...
973,768
sek788432/Waymo-2D-Object-Detection
resnet_ctl_imagenet_main.py
run
run
Run ResNet ImageNet training and eval loop using custom training loops.
[ "Run", "ResNet", "ImageNet", "training", "and", "eval", "loop", "using", "custom", "training", "loops." ]
def run(flags_obj): keras_utils.set_session_config() performance.set_mixed_precision_policy(flags_core.get_tf_dtype(flags_obj)) if tf.config.list_physical_devices('GPU'): if flags_obj.tf_gpu_thread_mode: keras_utils.set_gpu_thread_mode_and_count(per_gpu_thread_count=flags_obj.per_gpu_thr...
['def', 'run(flags_obj):', 'keras_utils.set_session_config()', 'performance.set_mixed_precision_policy(flags_core.get_tf_dtype(flags_obj))', 'if', "tf.config.list_physical_devices('GPU'):", 'if', 'flags_obj.tf_gpu_thread_mode:', 'keras_utils.set_gpu_thread_mode_and_count(per_gpu_thread_count=flags_obj.per_gpu_thread_co...
973,808
sek788432/Waymo-2D-Object-Detection
loss_utils.py
multi_level_flatten
multi_level_flatten
Flattens a multi-level input.
[ "Flattens", "a", "multi-level", "input." ]
def multi_level_flatten(multi_level_inputs, last_dim=None): flattened_inputs = [] batch_size = None for level in multi_level_inputs.keys(): single_input = multi_level_inputs[level] if batch_size is None: batch_size = single_input.shape[0] or tf.shape(single_input)[0] if l...
['def', 'multi_level_flatten(multi_level_inputs,', 'last_dim=None):', 'flattened_inputs', '=', '[]', 'batch_size', '=', 'None', 'for', 'level', 'in', 'multi_level_inputs.keys():', 'single_input', '=', 'multi_level_inputs[level]', 'if', 'batch_size', 'is', 'None:', 'batch_size', '=', 'single_input.shape[0]', 'or', 'tf.s...
973,813
sek788432/Waymo-2D-Object-Detection
anchor_generator.py
maybe_map_structure_for_anchor
maybe_map_structure_for_anchor
broadcast the params to match anchor_sizes.
[ "broadcast", "the", "params", "to", "match", "anchor_sizes." ]
def maybe_map_structure_for_anchor(params, anchor_sizes): if all((isinstance(param, (int, float)) for param in params)): if isinstance(anchor_sizes, (tuple, list)): return [params] * len(anchor_sizes) elif isinstance(anchor_sizes, dict): return tf.nest.map_structure(lambda _:...
['def', 'maybe_map_structure_for_anchor(params,', 'anchor_sizes):', 'if', 'all((isinstance(param,', '(int,', 'float))', 'for', 'param', 'in', 'params)):', 'if', 'isinstance(anchor_sizes,', '(tuple,', 'list)):', 'return', '[params]', '*', 'len(anchor_sizes)', 'elif', 'isinstance(anchor_sizes,', 'dict):', 'return', 'tf.n...
973,816
sek788432/Waymo-2D-Object-Detection
new_best_metric.py
NewBestMetric.metric_value
metric_value
Computes the metric value for the given `output`.
[ "Computes", "the", "metric", "value", "for", "the", "given", "`output`." ]
def metric_value(self, output: runner.Output) -> float: if callable(self.metric): value = self.metric(output) else: value = output[self.metric] return float(utils.get_value(value))
['def', 'metric_value(self,', 'output:', 'runner.Output)', '->', 'float:', 'if', 'callable(self.metric):', 'value', '=', 'self.metric(output)', 'else:', 'value', '=', 'output[self.metric]', 'return', 'float(utils.get_value(value))']
973,847
sek788432/Waymo-2D-Object-Detection
new_best_metric.py
NewBestMetric.best_value
best_value
Returns the best metric value seen so far.
[ "Returns", "the", "best", "metric", "value", "seen", "so", "far." ]
def best_value(self) -> float: return self._best_value.read()
['def', 'best_value(self)', '->', 'float:', 'return', 'self._best_value.read()']
973,848
sek788432/Waymo-2D-Object-Detection
new_best_metric.py
JSONPersistedValue.write
write
Writes the value, updating the backing store if one was provided.
[ "Writes", "the", "value,", "updating", "the", "backing", "store", "if", "one", "was", "provided." ]
def write(self, value): self._value = value if self._filename is not None and self._write_value: tmp_filename = f'{self._filename}.tmp.{uuid.uuid4().hex}' with tf.io.gfile.GFile(tmp_filename, 'w') as f: json.dump(self._value, f) tf.io.gfile.rename(tmp_filename, self._filename...
['def', 'write(self,', 'value):', 'self._value', '=', 'value', 'if', 'self._filename', 'is', 'not', 'None', 'and', 'self._write_value:', 'tmp_filename', '=', "f'{self._filename}.tmp.{uuid.uuid4().hex}'", 'with', 'tf.io.gfile.GFile(tmp_filename,', "'w')", 'as', 'f:', 'json.dump(self._value,', 'f)', 'tf.io.gfile.rename(t...
973,851
sek788432/Waymo-2D-Object-Detection
single_task_evaluator.py
SingleTaskEvaluator.eval_begin
eval_begin
Actions to take once before every eval loop.
[ "Actions", "to", "take", "once", "before", "every", "eval", "loop." ]
def eval_begin(self): for metric in self.metrics: metric.reset_states()
['def', 'eval_begin(self):', 'for', 'metric', 'in', 'self.metrics:', 'metric.reset_states()']
973,852
sek788432/Waymo-2D-Object-Detection
single_task_evaluator.py
SingleTaskEvaluator.eval_end
eval_end
Actions to take once after an eval loop.
[ "Actions", "to", "take", "once", "after", "an", "eval", "loop." ]
def eval_end(self): with self.strategy.scope(): metrics = {metric.name: metric.result() for metric in self.metrics} return metrics
['def', 'eval_end(self):', 'with', 'self.strategy.scope():', 'metrics', '=', '{metric.name:', 'metric.result()', 'for', 'metric', 'in', 'self.metrics}', 'return', 'metrics']
973,854
sek788432/Waymo-2D-Object-Detection
epoch_helper.py
EpochHelper.epoch_begin
epoch_begin
Returns whether a new epoch should begin.
[ "Returns", "whether", "a", "new", "epoch", "should", "begin." ]
def epoch_begin(self): if self._in_epoch: return False current_step = self._global_step.numpy() self._epoch_start_step = current_step self._current_epoch = current_step // self._epoch_steps self._in_epoch = True return True
['def', 'epoch_begin(self):', 'if', 'self._in_epoch:', 'return', 'False', 'current_step', '=', 'self._global_step.numpy()', 'self._epoch_start_step', '=', 'current_step', 'self._current_epoch', '=', 'current_step', '//', 'self._epoch_steps', 'self._in_epoch', '=', 'True', 'return', 'True']
973,861
sek788432/Waymo-2D-Object-Detection
model.py
axis_pad
axis_pad
Pad a tensor with the specified values along a single axis.
[ "Pad", "a", "tensor", "with", "the", "specified", "values", "along", "a", "single", "axis." ]
def axis_pad(tensor, axis, before=0, after=0, constant_values=0.0): if before == 0 and after == 0: return tensor ndims = tensor.shape.ndims padding_size = np.zeros((ndims, 2), dtype='int32') padding_size[axis] = (before, after) return tf.pad(tensor=tensor, paddings=tf.constant(padding_size),...
['def', 'axis_pad(tensor,', 'axis,', 'before=0,', 'after=0,', 'constant_values=0.0):', 'if', 'before', '==', '0', 'and', 'after', '==', '0:', 'return', 'tensor', 'ndims', '=', 'tensor.shape.ndims', 'padding_size', '=', 'np.zeros((ndims,', '2),', "dtype='int32')", 'padding_size[axis]', '=', '(before,', 'after)', 'return...
973,930
sek788432/Waymo-2D-Object-Detection
yamnet.py
class_names
class_names
Read the class name definition file and return a list of strings.
[ "Read", "the", "class", "name", "definition", "file", "and", "return", "a", "list", "of", "strings." ]
def class_names(class_map_csv): if tf.is_tensor(class_map_csv): class_map_csv = class_map_csv.numpy() with open(class_map_csv) as csv_file: reader = csv.reader(csv_file) next(reader) return np.array([display_name for (_, _, display_name) in reader])
['def', 'class_names(class_map_csv):', 'if', 'tf.is_tensor(class_map_csv):', 'class_map_csv', '=', 'class_map_csv.numpy()', 'with', 'open(class_map_csv)', 'as', 'csv_file:', 'reader', '=', 'csv.reader(csv_file)', 'next(reader)', 'return', 'np.array([display_name', 'for', '(_,', '_,', 'display_name)', 'in', 'reader])']
973,995
sek788432/Waymo-2D-Object-Detection
tasks.py
UnrolledTask.episode_batch
episode_batch
Returns a batch of episodes.
[ "Returns", "a", "batch", "of", "episodes." ]
def episode_batch(self, batch_size): batched_inputs = collections.OrderedDict([[mtype, []] for mtype in self.config.inputs]) batched_queries = [] batched_outputs = [] batched_masks = [] for _ in range(int(batch_size)): with self._lock: (inputs, query, outputs) = self.episode() ...
['def', 'episode_batch(self,', 'batch_size):', 'batched_inputs', '=', 'collections.OrderedDict([[mtype,', '[]]', 'for', 'mtype', 'in', 'self.config.inputs])', 'batched_queries', '=', '[]', 'batched_outputs', '=', '[]', 'batched_masks', '=', '[]', 'for', '_', 'in', 'range(int(batch_size)):', 'with', 'self._lock:', '(inp...
974,052
sek788432/Waymo-2D-Object-Detection
whiten.py
apply_whitening
apply_whitening
Applies the whitening to the descriptors as a post-processing step.
[ "Applies", "the", "whitening", "to", "the", "descriptors", "as", "a", "post-processing", "step." ]
def apply_whitening(descriptors, mean_descriptor_vector, projection, output_dim=None): eps = 1e-06 if output_dim is None: output_dim = projection.shape[0] descriptors = np.dot(projection[:output_dim, :], descriptors - mean_descriptor_vector) descriptors_whitened = descriptors / (np.linalg.norm(d...
['def', 'apply_whitening(descriptors,', 'mean_descriptor_vector,', 'projection,', 'output_dim=None):', 'eps', '=', '1e-06', 'if', 'output_dim', 'is', 'None:', 'output_dim', '=', 'projection.shape[0]', 'descriptors', '=', 'np.dot(projection[:output_dim,', ':],', 'descriptors', '-', 'mean_descriptor_vector)', 'descriptor...
974,248
sek788432/Waymo-2D-Object-Detection
utils.py
pil_imagenet_loader
pil_imagenet_loader
Pillow loader for the images.
[ "Pillow", "loader", "for", "the", "images." ]
def pil_imagenet_loader(path, imsize, bounding_box=None, preprocess=True): img = image_loading_utils.RgbLoader(path) if bounding_box is not None: imfullsize = max(img.size) img = img.crop(bounding_box) imsize = imsize * max(img.size) / imfullsize img.thumbnail((imsize, imsize), Image...
['def', 'pil_imagenet_loader(path,', 'imsize,', 'bounding_box=None,', 'preprocess=True):', 'img', '=', 'image_loading_utils.RgbLoader(path)', 'if', 'bounding_box', 'is', 'not', 'None:', 'imfullsize', '=', 'max(img.size)', 'img', '=', 'img.crop(bounding_box)', 'imsize', '=', 'imsize', '*', 'max(img.size)', '/', 'imfulls...
974,251
sek788432/Waymo-2D-Object-Detection
dataset_file_io.py
ReadSolution
ReadSolution
Reads solution from file, for a given task.
[ "Reads", "solution", "from", "file,", "for", "a", "given", "task." ]
def ReadSolution(file_path, task): public_solution = {} private_solution = {} ignored_ids = [] with tf.io.gfile.GFile(file_path, 'r') as csv_file: reader = csv.reader(csv_file) next(reader, None) for row in reader: test_id = row[0] if row[2] == 'Ignored': ...
['def', 'ReadSolution(file_path,', 'task):', 'public_solution', '=', '{}', 'private_solution', '=', '{}', 'ignored_ids', '=', '[]', 'with', 'tf.io.gfile.GFile(file_path,', "'r')", 'as', 'csv_file:', 'reader', '=', 'csv.reader(csv_file)', 'next(reader,', 'None)', 'for', 'row', 'in', 'reader:', 'test_id', '=', 'row[0]', ...
974,253
sek788432/Waymo-2D-Object-Detection
dataset.py
ReadMetricsFile
ReadMetricsFile
Reads aggregated retrieval metrics from text file.
[ "Reads", "aggregated", "retrieval", "metrics", "from", "text", "file." ]
def ReadMetricsFile(metrics_path): with tf.io.gfile.GFile(metrics_path, 'r') as f: file_contents_stripped = [l.rstrip() for l in f] if len(file_contents_stripped) % 4: raise ValueError('Malformed input %s: number of lines must be a multiple of 4, but it is %d' % (metrics_path, len(file_contents_...
['def', 'ReadMetricsFile(metrics_path):', 'with', 'tf.io.gfile.GFile(metrics_path,', "'r')", 'as', 'f:', 'file_contents_stripped', '=', '[l.rstrip()', 'for', 'l', 'in', 'f]', 'if', 'len(file_contents_stripped)', '%', '4:', 'raise', "ValueError('Malformed", 'input', '%s:', 'number', 'of', 'lines', 'must', 'be', 'a', 'mu...
974,269
sek788432/Waymo-2D-Object-Detection
dataset.py
CreateConfigForTestDataset
CreateConfigForTestDataset
Creates the configuration dictionary for the test dataset.
[ "Creates", "the", "configuration", "dictionary", "for", "the", "test", "dataset." ]
def CreateConfigForTestDataset(dataset, dir_main): dataset = dataset.lower() def _ConfigImname(cfg, i): return os.path.join(cfg['dir_images'], cfg['imlist'][i] + cfg['ext']) def _ConfigQimname(cfg, i): return os.path.join(cfg['dir_images'], cfg['qimlist'][i] + cfg['qext']) if dataset n...
['def', 'CreateConfigForTestDataset(dataset,', 'dir_main):', 'dataset', '=', 'dataset.lower()', 'def', '_ConfigImname(cfg,', 'i):', 'return', "os.path.join(cfg['dir_images'],", "cfg['imlist'][i]", '+', "cfg['ext'])", 'def', '_ConfigQimname(cfg,', 'i):', 'return', "os.path.join(cfg['dir_images'],", "cfg['qimlist'][i]", ...
974,270
sek788432/Waymo-2D-Object-Detection
normalization.py
L2Normalization.call
call
Invokes the L2Normalization instance.
[ "Invokes", "the", "L2Normalization", "instance." ]
def call(self, x, axis=1): return tf.nn.l2_normalize(x, axis, epsilon=self.eps)
['def', 'call(self,', 'x,', 'axis=1):', 'return', 'tf.nn.l2_normalize(x,', 'axis,', 'epsilon=self.eps)']
974,278
sek788432/Waymo-2D-Object-Detection
pooling.py
mac
mac
Performs global max pooling (MAC).
[ "Performs", "global", "max", "pooling", "(MAC)." ]
def mac(x, axis=None): if axis is None: axis = [1, 2] return tf.reduce_max(x, axis=axis, keepdims=False)
['def', 'mac(x,', 'axis=None):', 'if', 'axis', 'is', 'None:', 'axis', '=', '[1,', '2]', 'return', 'tf.reduce_max(x,', 'axis=axis,', 'keepdims=False)']
974,279
sek788432/Waymo-2D-Object-Detection
pooling.py
gem
gem
Performs generalized mean pooling (GeM).
[ "Performs", "generalized", "mean", "pooling", "(GeM)." ]
def gem(x, axis=None, power=3.0, eps=1e-06): if axis is None: axis = [1, 2] tmp = tf.pow(tf.maximum(x, eps), power) out = tf.pow(tf.reduce_mean(tmp, axis=axis, keepdims=False), 1.0 / power) return out
['def', 'gem(x,', 'axis=None,', 'power=3.0,', 'eps=1e-06):', 'if', 'axis', 'is', 'None:', 'axis', '=', '[1,', '2]', 'tmp', '=', 'tf.pow(tf.maximum(x,', 'eps),', 'power)', 'out', '=', 'tf.pow(tf.reduce_mean(tmp,', 'axis=axis,', 'keepdims=False),', '1.0', '/', 'power)', 'return', 'out']
974,281
sek788432/Waymo-2D-Object-Detection
pooling.py
MAC.call
call
Invokes the MAC pooling instance.
[ "Invokes", "the", "MAC", "pooling", "instance." ]
def call(self, x, axis=None): if axis is None: axis = [1, 2] return mac(x, axis=axis)
['def', 'call(self,', 'x,', 'axis=None):', 'if', 'axis', 'is', 'None:', 'axis', '=', '[1,', '2]', 'return', 'mac(x,', 'axis=axis)']
974,282
sek788432/Waymo-2D-Object-Detection
pooling.py
GeM.call
call
Invokes the GeM instance.
[ "Invokes", "the", "GeM", "instance." ]
def call(self, x, axis=None): if axis is None: axis = [1, 2] return gem(x, power=self.power, eps=self.eps, axis=axis)
['def', 'call(self,', 'x,', 'axis=None):', 'if', 'axis', 'is', 'None:', 'axis', '=', '[1,', '2]', 'return', 'gem(x,', 'power=self.power,', 'eps=self.eps,', 'axis=axis)']
974,284
sek788432/Waymo-2D-Object-Detection
global_features_utils.py
compute_metrics_and_print
compute_metrics_and_print
Computes and logs ground-truth metrics for Revisited datasets.
[ "Computes", "and", "logs", "ground-truth", "metrics", "for", "Revisited", "datasets." ]
def compute_metrics_and_print(dataset_name, sorted_index_ids, ground_truth, desired_pr_ranks=None, log=True): if dataset not in dataset.DATASET_NAMES: raise ValueError('Unknown dataset: {}!'.format(dataset)) if desired_pr_ranks is None: desired_pr_ranks = [1, 5, 10] (easy_ground_truth, mediu...
['def', 'compute_metrics_and_print(dataset_name,', 'sorted_index_ids,', 'ground_truth,', 'desired_pr_ranks=None,', 'log=True):', 'if', 'dataset', 'not', 'in', 'dataset.DATASET_NAMES:', 'raise', "ValueError('Unknown", 'dataset:', "{}!'.format(dataset))", 'if', 'desired_pr_ranks', 'is', 'None:', 'desired_pr_ranks', '=', ...
974,285
sek788432/Waymo-2D-Object-Detection
global_features_utils.py
AverageMeter.reset
reset
Resets all the values.
[ "Resets", "all", "the", "values." ]
def reset(self): self.val = 0 self.avg = 0 self.sum = 0 self.count = 0
['def', 'reset(self):', 'self.val', '=', '0', 'self.avg', '=', '0', 'self.sum', '=', '0', 'self.count', '=', '0']
974,290
sek788432/Waymo-2D-Object-Detection
delg_model.py
cosine_classifier_logits
cosine_classifier_logits
Compute cosine classifier logits using ArFace margin.
[ "Compute", "cosine", "classifier", "logits", "using", "ArFace", "margin." ]
def cosine_classifier_logits(prelogits, labels, num_classes, cosine_weights, scale_factor, arcface_margin, training=True): normalized_prelogits = tf.math.l2_normalize(prelogits, axis=1) normalized_weights = tf.math.l2_normalize(cosine_weights, axis=0) cosine_sim = tf.matmul(normalized_prelogits, normalized_...
['def', 'cosine_classifier_logits(prelogits,', 'labels,', 'num_classes,', 'cosine_weights,', 'scale_factor,', 'arcface_margin,', 'training=True):', 'normalized_prelogits', '=', 'tf.math.l2_normalize(prelogits,', 'axis=1)', 'normalized_weights', '=', 'tf.math.l2_normalize(cosine_weights,', 'axis=0)', 'cosine_sim', '=', ...
974,297
sek788432/Waymo-2D-Object-Detection
resnet50.py
ResNet50.call
call
Call the ResNet50 model.
[ "Call", "the", "ResNet50", "model." ]
def call(self, inputs, training=True, intermediates_dict=None): return self.build_call(inputs, training, intermediates_dict)
['def', 'call(self,', 'inputs,', 'training=True,', 'intermediates_dict=None):', 'return', 'self.build_call(inputs,', 'training,', 'intermediates_dict)']
974,304
sek788432/Waymo-2D-Object-Detection
agent.py
UvfAgentCore.clip_actions
clip_actions
Clip actions to spec.
[ "Clip", "actions", "to", "spec." ]
def clip_actions(self, actions): actions = tf.concat([tf.clip_by_value(actions[:, i:i + 1], self._action_spec.minimum[i], self._action_spec.maximum[i]) for i in range(self._action_spec.shape[0].value)], axis=1) return actions
['def', 'clip_actions(self,', 'actions):', 'actions', '=', 'tf.concat([tf.clip_by_value(actions[:,', 'i:i', '+', '1],', 'self._action_spec.minimum[i],', 'self._action_spec.maximum[i])', 'for', 'i', 'in', 'range(self._action_spec.shape[0].value)],', 'axis=1)', 'return', 'actions']
974,315
sek788432/Waymo-2D-Object-Detection
agent.py
UvfAgentCore.init_action_vars
init_action_vars
Create and return a tensorflow Variable holding an action.
[ "Create", "and", "return", "a", "tensorflow", "Variable", "holding", "an", "action." ]
def init_action_vars(self, name, i=None): if i is not None: name += '_%d' % i assert name not in self._action_vars, 'Conflict! %s is already initialized.' % name self._action_vars[name] = tf.Variable(self.sample_random_actions(1)[0], name='%s_action' % name) self._validate_actions(tf.expand_dims...
['def', 'init_action_vars(self,', 'name,', 'i=None):', 'if', 'i', 'is', 'not', 'None:', 'name', '+=', "'_%d'", '%', 'i', 'assert', 'name', 'not', 'in', 'self._action_vars,', "'Conflict!", '%s', 'is', 'already', "initialized.'", '%', 'name', 'self._action_vars[name]', '=', 'tf.Variable(self.sample_random_actions(1)[0],'...
974,320
sek788432/Waymo-2D-Object-Detection
cond_fn.py
true_fn
true_fn
Returns an op that evaluates to true.
[ "Returns", "an", "op", "that", "evaluates", "to", "true." ]
def true_fn(agent, state, action, transition_type, environment_steps, num_episodes): del agent, state, action, transition_type, environment_steps, num_episodes cond = tf.constant(True, dtype=tf.bool) return cond
['def', 'true_fn(agent,', 'state,', 'action,', 'transition_type,', 'environment_steps,', 'num_episodes):', 'del', 'agent,', 'state,', 'action,', 'transition_type,', 'environment_steps,', 'num_episodes', 'cond', '=', 'tf.constant(True,', 'dtype=tf.bool)', 'return', 'cond']
974,329
sek788432/Waymo-2D-Object-Detection
eval.py
get_eval_step
get_eval_step
Get one-step policy/env stepping ops.
[ "Get", "one-step", "policy/env", "stepping", "ops." ]
def get_eval_step(uvf_agent, state_preprocess, tf_env, action_fn, meta_action_fn, environment_steps, num_episodes, mode='eval'): tf_env.start_collect() state = tf_env.current_obs() action = action_fn(state, context=None) state_repr = state_preprocess(state) action_spec = tf_env.action_spec() act...
['def', 'get_eval_step(uvf_agent,', 'state_preprocess,', 'tf_env,', 'action_fn,', 'meta_action_fn,', 'environment_steps,', 'num_episodes,', "mode='eval'):", 'tf_env.start_collect()', 'state', '=', 'tf_env.current_obs()', 'action', '=', 'action_fn(state,', 'context=None)', 'state_repr', '=', 'state_preprocess(state)', '...
974,332
sek788432/Waymo-2D-Object-Detection
context.py
Context.create_vars
create_vars
Create tf variables for contexts.
[ "Create", "tf", "variables", "for", "contexts." ]
def create_vars(self, name, agent=None): if agent is not None: meta_vars = agent.create_vars(name) else: meta_vars = {} assert name not in self.context_vars, 'Conflict! %s is already initialized.' % name self.context_vars[name] = tuple([tf.Variable(tf.zeros(shape=spec.shape, dtype=spec.d...
['def', 'create_vars(self,', 'name,', 'agent=None):', 'if', 'agent', 'is', 'not', 'None:', 'meta_vars', '=', 'agent.create_vars(name)', 'else:', 'meta_vars', '=', '{}', 'assert', 'name', 'not', 'in', 'self.context_vars,', "'Conflict!", '%s', 'is', 'already', "initialized.'", '%', 'name', 'self.context_vars[name]', '=',...
974,380
sek788432/Waymo-2D-Object-Detection
rewards_functions.py
ctrl_rewards
ctrl_rewards
Returns the negative control cost.
[ "Returns", "the", "negative", "control", "cost." ]
def ctrl_rewards(states, actions, rewards, next_states, contexts, reward_scales=1.0): del states, rewards, contexts if actions is None: rewards = tf.to_float(tf.zeros(shape=next_states.shape[:1])) else: rewards = -tf.reduce_sum(tf.square(actions), axis=1) rewards *= reward_scales ...
['def', 'ctrl_rewards(states,', 'actions,', 'rewards,', 'next_states,', 'contexts,', 'reward_scales=1.0):', 'del', 'states,', 'rewards,', 'contexts', 'if', 'actions', 'is', 'None:', 'rewards', '=', 'tf.to_float(tf.zeros(shape=next_states.shape[:1]))', 'else:', 'rewards', '=', '-tf.reduce_sum(tf.square(actions),', 'axis...
974,396
sek788432/Waymo-2D-Object-Detection
utils.py
get_contextual_env_base
get_contextual_env_base
Wrap env_base with additional tf ops.
[ "Wrap", "env_base", "with", "additional", "tf", "ops." ]
def get_contextual_env_base(env_base, begin_ops=None, end_ops=None): def init(self_, env_base): self_._env_base = env_base attribute_list = ['_render_mode', '_gym_env'] for attribute in attribute_list: if hasattr(env_base, attribute): setattr(self_, attribute, ge...
['def', 'get_contextual_env_base(env_base,', 'begin_ops=None,', 'end_ops=None):', 'def', 'init(self_,', 'env_base):', 'self_._env_base', '=', 'env_base', 'attribute_list', '=', "['_render_mode',", "'_gym_env']", 'for', 'attribute', 'in', 'attribute_list:', 'if', 'hasattr(env_base,', 'attribute):', 'setattr(self_,', 'at...
974,405
sek788432/Waymo-2D-Object-Detection
utils.py
identity_vars
identity_vars
Return the identity ops for a list of tensors.
[ "Return", "the", "identity", "ops", "for", "a", "list", "of", "tensors." ]
def identity_vars(vars_): return [tf.identity(var) for var in vars_]
['def', 'identity_vars(vars_):', 'return', '[tf.identity(var)', 'for', 'var', 'in', 'vars_]']
974,410
sek788432/Waymo-2D-Object-Detection
tf_sequence_example_decoder.py
TFSequenceExampleDecoderHelper.decode
decode
Decodes the given serialized TF-SequenceExample.
[ "Decodes", "the", "given", "serialized", "TF-SequenceExample." ]
def decode(self, serialized_example, items=None): (context, feature_list) = tf.parse_single_sequence_example(serialized_example, self._keys_to_context_features, self._keys_to_sequence_features) for k in self._keys_to_context_features: v = self._keys_to_context_features[k] if isinstance(v, tf.Fix...
['def', 'decode(self,', 'serialized_example,', 'items=None):', '(context,', 'feature_list)', '=', 'tf.parse_single_sequence_example(serialized_example,', 'self._keys_to_context_features,', 'self._keys_to_sequence_features)', 'for', 'k', 'in', 'self._keys_to_context_features:', 'v', '=', 'self._keys_to_context_features[...
974,489
sek788432/Waymo-2D-Object-Detection
utils.py
quantize_op
quantize_op
Inserts a fake quantization op after inputs.
[ "Inserts", "a", "fake", "quantization", "op", "after", "inputs." ]
def quantize_op(inputs, is_training=True, is_quantized=True, default_min=0, default_max=6, ema_decay=0.999, scope='quant'): if not is_quantized: return inputs with tf.variable_scope(scope): min_var = _quant_var('min', default_min) max_var = _quant_var('max', default_max) if not i...
['def', 'quantize_op(inputs,', 'is_training=True,', 'is_quantized=True,', 'default_min=0,', 'default_max=6,', 'ema_decay=0.999,', "scope='quant'):", 'if', 'not', 'is_quantized:', 'return', 'inputs', 'with', 'tf.variable_scope(scope):', 'min_var', '=', "_quant_var('min',", 'default_min)', 'max_var', '=', "_quant_var('ma...
974,502
sek788432/Waymo-2D-Object-Detection
export_tflite_graph_lib_tf2.py
CenterNetModule.inference_fn
inference_fn
Encapsulates CenterNet inference for TFLite conversion.
[ "Encapsulates", "CenterNet", "inference", "for", "TFLite", "conversion." ]
def inference_fn(self, image): image = tf.cast(image, tf.float32) (image, shapes) = self._model.preprocess(image) prediction_dict = self._model.predict(image, None) detections = self._model.postprocess(prediction_dict, true_image_shapes=shapes) field_names = fields.DetectionResultFields classes_...
['def', 'inference_fn(self,', 'image):', 'image', '=', 'tf.cast(image,', 'tf.float32)', '(image,', 'shapes)', '=', 'self._model.preprocess(image)', 'prediction_dict', '=', 'self._model.predict(image,', 'None)', 'detections', '=', 'self._model.postprocess(prediction_dict,', 'true_image_shapes=shapes)', 'field_names', '=...
974,557
sek788432/Waymo-2D-Object-Detection
model_lib.py
continuous_eval
continuous_eval
Performs continuous evaluation on checkpoints written to a model directory.
[ "Performs", "continuous", "evaluation", "on", "checkpoints", "written", "to", "a", "model", "directory." ]
def continuous_eval(estimator, model_dir, input_fn, train_steps, name, max_retries=0): for (current_step, eval_results) in continuous_eval_generator(estimator, model_dir, input_fn, train_steps, name, max_retries): tf.logging.info('Step %s, Eval results: %s', current_step, eval_results)
['def', 'continuous_eval(estimator,', 'model_dir,', 'input_fn,', 'train_steps,', 'name,', 'max_retries=0):', 'for', '(current_step,', 'eval_results)', 'in', 'continuous_eval_generator(estimator,', 'model_dir,', 'input_fn,', 'train_steps,', 'name,', 'max_retries):', "tf.logging.info('Step", '%s,', 'Eval', 'results:', "%...
974,599
sek788432/Waymo-2D-Object-Detection
model_lib_tf1_test.py
ModelLibTest.test_model_fn_in_keypoints_eval_mode
test_model_fn_in_keypoints_eval_mode
Tests the model function in EVAL mode with keypoints config.
[ "Tests", "the", "model", "function", "in", "EVAL", "mode", "with", "keypoints", "config." ]
def test_model_fn_in_keypoints_eval_mode(self): configs = _get_configs_for_model(MODEL_NAME_FOR_KEYPOINTS_TEST) estimator_spec = self._assert_model_fn_for_train_eval(configs, 'eval') metric_ops = estimator_spec.eval_metric_ops self.assertIn('Keypoints_Precision/mAP ByCategory/face', metric_ops) self...
['def', 'test_model_fn_in_keypoints_eval_mode(self):', 'configs', '=', '_get_configs_for_model(MODEL_NAME_FOR_KEYPOINTS_TEST)', 'estimator_spec', '=', 'self._assert_model_fn_for_train_eval(configs,', "'eval')", 'metric_ops', '=', 'estimator_spec.eval_metric_ops', "self.assertIn('Keypoints_Precision/mAP", "ByCategory/fa...
974,609
sek788432/Waymo-2D-Object-Detection
model_builder_tf2_test.py
ModelBuilderTF2Test.test_create_center_net_model_mobilenet
test_create_center_net_model_mobilenet
Test building a CenterNet model using bilinear interpolation.
[ "Test", "building", "a", "CenterNet", "model", "using", "bilinear", "interpolation." ]
def test_create_center_net_model_mobilenet(self): proto_txt = '\n center_net {\n num_classes: 10\n feature_extractor {\n type: "mobilenet_v2_fpn"\n depth_multiplier: 1.0\n use_separable_conv: true\n upsampling_interpolation: "bilinear"\n }\n image...
['def', 'test_create_center_net_model_mobilenet(self):', 'proto_txt', '=', "'\\n", 'center_net', '{\\n', 'num_classes:', '10\\n', 'feature_extractor', '{\\n', 'type:', '"mobilenet_v2_fpn"\\n', 'depth_multiplier:', '1.0\\n', 'use_separable_conv:', 'true\\n', 'upsampling_interpolation:', '"bilinear"\\n', '}\\n', 'image_r...
974,701
sek788432/Waymo-2D-Object-Detection
densepose_ops.py
DensePoseHorizontalFlip.flip_parts_and_coords
flip_parts_and_coords
Flips part ids and coordinates.
[ "Flips", "part", "ids", "and", "coordinates." ]
def flip_parts_and_coords(self, part_ids, vu): (num_instances, num_points) = shape_utils.combined_static_and_dynamic_shape(part_ids) part_ids_flattened = tf.reshape(part_ids, [-1]) new_part_ids_flattened = tf.gather(self.part_symmetries, part_ids_flattened) new_part_ids = tf.reshape(new_part_ids_flatten...
['def', 'flip_parts_and_coords(self,', 'part_ids,', 'vu):', '(num_instances,', 'num_points)', '=', 'shape_utils.combined_static_and_dynamic_shape(part_ids)', 'part_ids_flattened', '=', 'tf.reshape(part_ids,', '[-1])', 'new_part_ids_flattened', '=', 'tf.gather(self.part_symmetries,', 'part_ids_flattened)', 'new_part_ids...
974,778
sek788432/Waymo-2D-Object-Detection
preprocessor.py
random_jitter_boxes
random_jitter_boxes
Randomly jitters boxes in image.
[ "Randomly", "jitters", "boxes", "in", "image." ]
def random_jitter_boxes(boxes, ratio=0.05, jitter_mode='default', seed=None): with tf.name_scope('RandomJitterBoxes'): (ymin, xmin, ymax, xmax) = (boxes[:, i] for i in range(4)) blist = box_list.BoxList(boxes) (ycenter, xcenter, height, width) = blist.get_center_coordinates_and_sizes() ...
['def', 'random_jitter_boxes(boxes,', 'ratio=0.05,', "jitter_mode='default',", 'seed=None):', 'with', "tf.name_scope('RandomJitterBoxes'):", '(ymin,', 'xmin,', 'ymax,', 'xmax)', '=', '(boxes[:,', 'i]', 'for', 'i', 'in', 'range(4))', 'blist', '=', 'box_list.BoxList(boxes)', '(ycenter,', 'xcenter,', 'height,', 'width)', ...
974,839
sek788432/Waymo-2D-Object-Detection
target_assigner_test.py
CenterNetCenterHeatmapTargetAssignerTest.test_weights
test_weights
Test that the weights correctly ignore ground truth.
[ "Test", "that", "the", "weights", "correctly", "ignore", "ground", "truth." ]
def test_weights(self): def graph1_fn(): box_batch = [tf.constant([self._box_center, self._box_lower_left]), tf.constant([self._box_center]), tf.constant([self._box_center_small])] classes = [tf.one_hot([0, 1], depth=4), tf.one_hot([2], depth=4), tf.one_hot([3], depth=4)] assigner = targeta...
['def', 'test_weights(self):', 'def', 'graph1_fn():', 'box_batch', '=', '[tf.constant([self._box_center,', 'self._box_lower_left]),', 'tf.constant([self._box_center]),', 'tf.constant([self._box_center_small])]', 'classes', '=', '[tf.one_hot([0,', '1],', 'depth=4),', 'tf.one_hot([2],', 'depth=4),', 'tf.one_hot([3],', 'd...
974,924
sek788432/Waymo-2D-Object-Detection
target_assigner_test.py
CornerOffsetTargetAssignerTest.test_filter_overlap_min_area_empty
test_filter_overlap_min_area_empty
Test that empty masks work on CPU.
[ "Test", "that", "empty", "masks", "work", "on", "CPU." ]
def test_filter_overlap_min_area_empty(self): def graph_fn(masks): return targetassigner.filter_mask_overlap_min_area(masks) masks = self.execute_cpu(graph_fn, [np.zeros((0, 5, 5), dtype=np.float32)]) self.assertEqual(masks.shape, (0, 5, 5))
['def', 'test_filter_overlap_min_area_empty(self):', 'def', 'graph_fn(masks):', 'return', 'targetassigner.filter_mask_overlap_min_area(masks)', 'masks', '=', 'self.execute_cpu(graph_fn,', '[np.zeros((0,', '5,', '5),', 'dtype=np.float32)])', 'self.assertEqual(masks.shape,', '(0,', '5,', '5))']
974,934
sek788432/Waymo-2D-Object-Detection
oid_hierarchical_labels_expansion.py
OIDHierarchicalLabelsExpansion.expand_labels_from_csv
expand_labels_from_csv
Expands a row containing labels from CSV file.
[ "Expands", "a", "row", "containing", "labels", "from", "CSV", "file." ]
def expand_labels_from_csv(self, csv_row, labelname_column_index=1, confidence_column_index=2): split_csv_row = six.ensure_str(csv_row).split(',') result = [csv_row] if int(split_csv_row[confidence_column_index]) == 1: assert split_csv_row[labelname_column_index] in self._hierarchy_keyed_child ...
['def', 'expand_labels_from_csv(self,', 'csv_row,', 'labelname_column_index=1,', 'confidence_column_index=2):', 'split_csv_row', '=', "six.ensure_str(csv_row).split(',')", 'result', '=', '[csv_row]', 'if', 'int(split_csv_row[confidence_column_index])', '==', '1:', 'assert', 'split_csv_row[labelname_column_index]', 'in'...
974,954
sek788432/Waymo-2D-Object-Detection
seq_example_util.py
sequence_bytes_feature
sequence_bytes_feature
Converts a bytes float array to a sequence bytes feature.
[ "Converts", "a", "bytes", "float", "array", "to", "a", "sequence", "bytes", "feature." ]
def sequence_bytes_feature(ndarray): feature_list = tf.train.FeatureList() for row in ndarray: if isinstance(row, np.ndarray): row = row.tolist() feature = feature_list.feature.add() if row: row = [tf.compat.as_bytes(val) for val in row] feature.bytes_...
['def', 'sequence_bytes_feature(ndarray):', 'feature_list', '=', 'tf.train.FeatureList()', 'for', 'row', 'in', 'ndarray:', 'if', 'isinstance(row,', 'np.ndarray):', 'row', '=', 'row.tolist()', 'feature', '=', 'feature_list.feature.add()', 'if', 'row:', 'row', '=', '[tf.compat.as_bytes(val)', 'for', 'val', 'in', 'row]', ...
974,961
sek788432/Waymo-2D-Object-Detection
center_net_meta_arch.py
argmax_feature_map_locations
argmax_feature_map_locations
Returns the peak locations in the feature map.
[ "Returns", "the", "peak", "locations", "in", "the", "feature", "map." ]
def argmax_feature_map_locations(feature_map): (batch_size, _, width, num_channels) = _get_shape(feature_map, 4) feature_map_flattened = tf.reshape(feature_map, [batch_size, -1, num_channels]) peak_flat_indices = tf.math.argmax(feature_map_flattened, axis=1, output_type=tf.dtypes.int32) (y_indices, x_in...
['def', 'argmax_feature_map_locations(feature_map):', '(batch_size,', '_,', 'width,', 'num_channels)', '=', '_get_shape(feature_map,', '4)', 'feature_map_flattened', '=', 'tf.reshape(feature_map,', '[batch_size,', '-1,', 'num_channels])', 'peak_flat_indices', '=', 'tf.math.argmax(feature_map_flattened,', 'axis=1,', 'ou...
974,995
sek788432/Waymo-2D-Object-Detection
center_net_meta_arch.py
row_col_indices_from_flattened_indices
row_col_indices_from_flattened_indices
Computes row and column indices from flattened indices.
[ "Computes", "row", "and", "column", "indices", "from", "flattened", "indices." ]
def row_col_indices_from_flattened_indices(indices, num_cols): row_indices = indices // num_cols col_indices = indices - row_indices * num_cols return (row_indices, col_indices)
['def', 'row_col_indices_from_flattened_indices(indices,', 'num_cols):', 'row_indices', '=', 'indices', '//', 'num_cols', 'col_indices', '=', 'indices', '-', 'row_indices', '*', 'num_cols', 'return', '(row_indices,', 'col_indices)']
975,000
sek788432/Waymo-2D-Object-Detection
center_net_meta_arch_tf2_test.py
get_fake_center_params
get_fake_center_params
Returns the fake object center parameter namedtuple.
[ "Returns", "the", "fake", "object", "center", "parameter", "namedtuple." ]
def get_fake_center_params(max_box_predictions=5): return cnma.ObjectCenterParams(classification_loss=losses.WeightedSigmoidClassificationLoss(), object_center_loss_weight=1.0, min_box_overlap_iou=1.0, max_box_predictions=max_box_predictions, use_labeled_classes=False, center_head_num_filters=[128], center_head_ker...
['def', 'get_fake_center_params(max_box_predictions=5):', 'return', 'cnma.ObjectCenterParams(classification_loss=losses.WeightedSigmoidClassificationLoss(),', 'object_center_loss_weight=1.0,', 'min_box_overlap_iou=1.0,', 'max_box_predictions=max_box_predictions,', 'use_labeled_classes=False,', 'center_head_num_filters=...
975,016
sek788432/Waymo-2D-Object-Detection
center_net_meta_arch_tf2_test.py
CenterNetMetaArchTest.test_loss
test_loss
Test the loss function.
[ "Test", "the", "loss", "function." ]
def test_loss(self): groundtruth_dict = get_fake_groundtruth_dict(16, 32, 4) model = build_center_net_meta_arch() model.provide_groundtruth(groundtruth_boxes_list=groundtruth_dict[fields.BoxListFields.boxes], groundtruth_weights_list=groundtruth_dict[fields.BoxListFields.weights], groundtruth_classes_list=g...
['def', 'test_loss(self):', 'groundtruth_dict', '=', 'get_fake_groundtruth_dict(16,', '32,', '4)', 'model', '=', 'build_center_net_meta_arch()', 'model.provide_groundtruth(groundtruth_boxes_list=groundtruth_dict[fields.BoxListFields.boxes],', 'groundtruth_weights_list=groundtruth_dict[fields.BoxListFields.weights],', '...
975,032
sek788432/Waymo-2D-Object-Detection
center_net_meta_arch_tf2_test.py
CenterNetMetaArchRestoreTest.test_retore_map_detection
test_retore_map_detection
Test that detection checkpoints can be restored.
[ "Test", "that", "detection", "checkpoints", "can", "be", "restored." ]
def test_retore_map_detection(self): model = build_center_net_meta_arch(build_resnet=True) restore_from_objects_map = model.restore_from_objects('detection') self.assertIsInstance(restore_from_objects_map['model']._feature_extractor, tf.keras.Model)
['def', 'test_retore_map_detection(self):', 'model', '=', 'build_center_net_meta_arch(build_resnet=True)', 'restore_from_objects_map', '=', "model.restore_from_objects('detection')", "self.assertIsInstance(restore_from_objects_map['model']._feature_extractor,", 'tf.keras.Model)']
975,040
sek788432/Waymo-2D-Object-Detection
deepmac_meta_arch.py
crop_masks_within_boxes
crop_masks_within_boxes
Crops masks to lie tightly within the boxes.
[ "Crops", "masks", "to", "lie", "tightly", "within", "the", "boxes." ]
def crop_masks_within_boxes(masks, boxes, output_size): masks = spatial_transform_ops.matmul_crop_and_resize(masks[:, :, :, tf.newaxis], boxes[:, tf.newaxis, :], [output_size, output_size]) return masks[:, 0, :, :, 0]
['def', 'crop_masks_within_boxes(masks,', 'boxes,', 'output_size):', 'masks', '=', 'spatial_transform_ops.matmul_crop_and_resize(masks[:,', ':,', ':,', 'tf.newaxis],', 'boxes[:,', 'tf.newaxis,', ':],', '[output_size,', 'output_size])', 'return', 'masks[:,', '0,', ':,', ':,', '0]']
975,055
sek788432/Waymo-2D-Object-Detection
deepmac_meta_arch.py
deepmac_proto_to_params
deepmac_proto_to_params
Convert proto to named tuple.
[ "Convert", "proto", "to", "named", "tuple." ]
def deepmac_proto_to_params(deepmac_config): loss = losses_pb2.Loss() loss.localization_loss.weighted_l2.CopyFrom(losses_pb2.WeightedL2LocalizationLoss()) loss.classification_loss.CopyFrom(deepmac_config.classification_loss) (classification_loss, _, _, _, _, _, _) = losses_builder.build(loss) jitter...
['def', 'deepmac_proto_to_params(deepmac_config):', 'loss', '=', 'losses_pb2.Loss()', 'loss.localization_loss.weighted_l2.CopyFrom(losses_pb2.WeightedL2LocalizationLoss())', 'loss.classification_loss.CopyFrom(deepmac_config.classification_loss)', '(classification_loss,', '_,', '_,', '_,', '_,', '_,', '_)', '=', 'losses...
975,057
sek788432/Waymo-2D-Object-Detection
deepmac_meta_arch.py
DeepMACMetaArch.postprocess
postprocess
Produces boxes given a prediction dict returned by predict().
[ "Produces", "boxes", "given", "a", "prediction", "dict", "returned", "by", "predict()." ]
def postprocess(self, prediction_dict, true_image_shapes, **params): postprocess_dict = super(DeepMACMetaArch, self).postprocess(prediction_dict, true_image_shapes, **params) boxes_strided = postprocess_dict['detection_boxes_strided'] if self._deepmac_params is not None: masks = self._postprocess_ma...
['def', 'postprocess(self,', 'prediction_dict,', 'true_image_shapes,', '**params):', 'postprocess_dict', '=', 'super(DeepMACMetaArch,', 'self).postprocess(prediction_dict,', 'true_image_shapes,', '**params)', 'boxes_strided', '=', "postprocess_dict['detection_boxes_strided']", 'if', 'self._deepmac_params', 'is', 'not',...
975,058
sek788432/Waymo-2D-Object-Detection
deepmac_meta_arch.py
DeepMACMetaArch.predict_masks_from_boxes
predict_masks_from_boxes
Produces masks for the provided boxes.
[ "Produces", "masks", "for", "the", "provided", "boxes." ]
def predict_masks_from_boxes(self, prediction_dict, true_image_shapes, provided_boxes, **params): postprocess_dict = super(DeepMACMetaArch, self).postprocess(prediction_dict, true_image_shapes, **params) instance_embedding = prediction_dict[INSTANCE_EMBEDDING][-1] resized_image_shapes = shape_utils.combined...
['def', 'predict_masks_from_boxes(self,', 'prediction_dict,', 'true_image_shapes,', 'provided_boxes,', '**params):', 'postprocess_dict', '=', 'super(DeepMACMetaArch,', 'self).postprocess(prediction_dict,', 'true_image_shapes,', '**params)', 'instance_embedding', '=', 'prediction_dict[INSTANCE_EMBEDDING][-1]', 'resized_...
975,059
sek788432/Waymo-2D-Object-Detection
deepmac_meta_arch_test.py
build_meta_arch
build_meta_arch
Builds the DeepMAC meta architecture.
[ "Builds", "the", "DeepMAC", "meta", "architecture." ]
def build_meta_arch(predict_full_resolution_masks=False, use_dice_loss=False): feature_extractor = DummyFeatureExtractor(channel_means=(1.0, 2.0, 3.0), channel_stds=(10.0, 20.0, 30.0), bgr_ordering=False, num_feature_outputs=2, stride=4) image_resizer_fn = functools.partial(preprocessor.resize_to_range, min_dim...
['def', 'build_meta_arch(predict_full_resolution_masks=False,', 'use_dice_loss=False):', 'feature_extractor', '=', 'DummyFeatureExtractor(channel_means=(1.0,', '2.0,', '3.0),', 'channel_stds=(10.0,', '20.0,', '30.0),', 'bgr_ordering=False,', 'num_feature_outputs=2,', 'stride=4)', 'image_resizer_fn', '=', 'functools.par...
975,060
sek788432/Waymo-2D-Object-Detection
center_net_hourglass_feature_extractor.py
hourglass_10
hourglass_10
The Hourglass-10 backbone for CenterNet.
[ "The", "Hourglass-10", "backbone", "for", "CenterNet." ]
def hourglass_10(channel_means, channel_stds, bgr_ordering, **kwargs): del kwargs network = hourglass_network.hourglass_10(num_channels=32) return CenterNetHourglassFeatureExtractor(network, channel_means=channel_means, channel_stds=channel_stds, bgr_ordering=bgr_ordering)
['def', 'hourglass_10(channel_means,', 'channel_stds,', 'bgr_ordering,', '**kwargs):', 'del', 'kwargs', 'network', '=', 'hourglass_network.hourglass_10(num_channels=32)', 'return', 'CenterNetHourglassFeatureExtractor(network,', 'channel_means=channel_means,', 'channel_stds=channel_stds,', 'bgr_ordering=bgr_ordering)']
975,171
sek788432/Waymo-2D-Object-Detection
center_net_hourglass_feature_extractor.py
hourglass_20
hourglass_20
The Hourglass-20 backbone for CenterNet.
[ "The", "Hourglass-20", "backbone", "for", "CenterNet." ]
def hourglass_20(channel_means, channel_stds, bgr_ordering, **kwargs): del kwargs network = hourglass_network.hourglass_20(num_channels=48) return CenterNetHourglassFeatureExtractor(network, channel_means=channel_means, channel_stds=channel_stds, bgr_ordering=bgr_ordering)
['def', 'hourglass_20(channel_means,', 'channel_stds,', 'bgr_ordering,', '**kwargs):', 'del', 'kwargs', 'network', '=', 'hourglass_network.hourglass_20(num_channels=48)', 'return', 'CenterNetHourglassFeatureExtractor(network,', 'channel_means=channel_means,', 'channel_stds=channel_stds,', 'bgr_ordering=bgr_ordering)']
975,172
sek788432/Waymo-2D-Object-Detection
center_net_hourglass_feature_extractor.py
hourglass_32
hourglass_32
The Hourglass-32 backbone for CenterNet.
[ "The", "Hourglass-32", "backbone", "for", "CenterNet." ]
def hourglass_32(channel_means, channel_stds, bgr_ordering, **kwargs): del kwargs network = hourglass_network.hourglass_32(num_channels=48) return CenterNetHourglassFeatureExtractor(network, channel_means=channel_means, channel_stds=channel_stds, bgr_ordering=bgr_ordering)
['def', 'hourglass_32(channel_means,', 'channel_stds,', 'bgr_ordering,', '**kwargs):', 'del', 'kwargs', 'network', '=', 'hourglass_network.hourglass_32(num_channels=48)', 'return', 'CenterNetHourglassFeatureExtractor(network,', 'channel_means=channel_means,', 'channel_stds=channel_stds,', 'bgr_ordering=bgr_ordering)']
975,173
sek788432/Waymo-2D-Object-Detection
center_net_hourglass_feature_extractor.py
hourglass_52
hourglass_52
The Hourglass-52 backbone for CenterNet.
[ "The", "Hourglass-52", "backbone", "for", "CenterNet." ]
def hourglass_52(channel_means, channel_stds, bgr_ordering, **kwargs): del kwargs network = hourglass_network.hourglass_52(num_channels=64) return CenterNetHourglassFeatureExtractor(network, channel_means=channel_means, channel_stds=channel_stds, bgr_ordering=bgr_ordering)
['def', 'hourglass_52(channel_means,', 'channel_stds,', 'bgr_ordering,', '**kwargs):', 'del', 'kwargs', 'network', '=', 'hourglass_network.hourglass_52(num_channels=64)', 'return', 'CenterNetHourglassFeatureExtractor(network,', 'channel_means=channel_means,', 'channel_stds=channel_stds,', 'bgr_ordering=bgr_ordering)']
975,174
sek788432/Waymo-2D-Object-Detection
center_net_mobilenet_v2_feature_extractor.py
mobilenet_v2
mobilenet_v2
The MobileNetV2 backbone for CenterNet.
[ "The", "MobileNetV2", "backbone", "for", "CenterNet." ]
def mobilenet_v2(channel_means, channel_stds, bgr_ordering, depth_multiplier=1.0, **kwargs): del kwargs network = mobilenetv2.mobilenet_v2(batchnorm_training=True, alpha=depth_multiplier, include_top=False, weights='imagenet' if depth_multiplier == 1.0 else None) return CenterNetMobileNetV2FeatureExtractor(...
['def', 'mobilenet_v2(channel_means,', 'channel_stds,', 'bgr_ordering,', 'depth_multiplier=1.0,', '**kwargs):', 'del', 'kwargs', 'network', '=', 'mobilenetv2.mobilenet_v2(batchnorm_training=True,', 'alpha=depth_multiplier,', 'include_top=False,', "weights='imagenet'", 'if', 'depth_multiplier', '==', '1.0', 'else', 'Non...
975,178
sek788432/Waymo-2D-Object-Detection
center_net_resnet_feature_extractor.py
resnet_v2_101
resnet_v2_101
The ResNet v2 101 feature extractor.
[ "The", "ResNet", "v2", "101", "feature", "extractor." ]
def resnet_v2_101(channel_means, channel_stds, bgr_ordering, **kwargs): del kwargs return CenterNetResnetFeatureExtractor(resnet_type='resnet_v2_101', channel_means=channel_means, channel_stds=channel_stds, bgr_ordering=bgr_ordering)
['def', 'resnet_v2_101(channel_means,', 'channel_stds,', 'bgr_ordering,', '**kwargs):', 'del', 'kwargs', 'return', "CenterNetResnetFeatureExtractor(resnet_type='resnet_v2_101',", 'channel_means=channel_means,', 'channel_stds=channel_stds,', 'bgr_ordering=bgr_ordering)']
975,184
sek788432/Waymo-2D-Object-Detection
ssd_mobiledet_feature_extractor.py
mobiledet_gpu_backbone
mobiledet_gpu_backbone
Build a MobileDet GPU backbone.
[ "Build", "a", "MobileDet", "GPU", "backbone." ]
def mobiledet_gpu_backbone(h, multiplier=1.0): def _scale(filters): return _scale_filters(filters, multiplier) ibn = functools.partial(_inverted_bottleneck, activation_fn=tf.nn.relu6) fused = functools.partial(_fused_conv, activation_fn=tf.nn.relu6) tucker = functools.partial(_tucker_conv, acti...
['def', 'mobiledet_gpu_backbone(h,', 'multiplier=1.0):', 'def', '_scale(filters):', 'return', '_scale_filters(filters,', 'multiplier)', 'ibn', '=', 'functools.partial(_inverted_bottleneck,', 'activation_fn=tf.nn.relu6)', 'fused', '=', 'functools.partial(_fused_conv,', 'activation_fn=tf.nn.relu6)', 'tucker', '=', 'funct...
975,234
sek788432/Waymo-2D-Object-Detection
resnet_v1.py
block_basic
block_basic
A residual block for ResNet18/34.
[ "A", "residual", "block", "for", "ResNet18/34." ]
def block_basic(x, filters, kernel_size=3, stride=1, conv_shortcut=False, name=None): layers = tf.keras.layers bn_axis = 3 if tf.keras.backend.image_data_format() == 'channels_last' else 1 preact = layers.BatchNormalization(axis=bn_axis, epsilon=1.001e-05, name=name + '_preact_bn')(x) preact = layers.Ac...
['def', 'block_basic(x,', 'filters,', 'kernel_size=3,', 'stride=1,', 'conv_shortcut=False,', 'name=None):', 'layers', '=', 'tf.keras.layers', 'bn_axis', '=', '3', 'if', 'tf.keras.backend.image_data_format()', '==', "'channels_last'", 'else', '1', 'preact', '=', 'layers.BatchNormalization(axis=bn_axis,', 'epsilon=1.001e...
975,294
sek788432/Waymo-2D-Object-Detection
bifpn_utils.py
create_downsample_feature_map_ops
create_downsample_feature_map_ops
Creates Keras layers for downsampling feature maps.
[ "Creates", "Keras", "layers", "for", "downsampling", "feature", "maps." ]
def create_downsample_feature_map_ops(scale, downsample_method, conv_hyperparams, is_training, freeze_batchnorm, name): layers = [] padding = 'SAME' stride = int(scale) kernel_size = stride + 1 if downsample_method == 'max_pooling': layers.append(tf.keras.layers.MaxPooling2D(pool_size=kernel...
['def', 'create_downsample_feature_map_ops(scale,', 'downsample_method,', 'conv_hyperparams,', 'is_training,', 'freeze_batchnorm,', 'name):', 'layers', '=', '[]', 'padding', '=', "'SAME'", 'stride', '=', 'int(scale)', 'kernel_size', '=', 'stride', '+', '1', 'if', 'downsample_method', '==', "'max_pooling':", 'layers.app...
975,362
sek788432/Waymo-2D-Object-Detection
config_util_test.py
ConfigUtilTest.testOverwriteSampleFromDatasetWeights
testOverwriteSampleFromDatasetWeights
Tests config override for sample_from_datasets_weights.
[ "Tests", "config", "override", "for", "sample_from_datasets_weights." ]
def testOverwriteSampleFromDatasetWeights(self): pipeline_config = pipeline_pb2.TrainEvalPipelineConfig() pipeline_config.train_input_reader.sample_from_datasets_weights.extend([1, 2]) pipeline_config_path = os.path.join(self.get_temp_dir(), 'pipeline.config') _write_config(pipeline_config, pipeline_con...
['def', 'testOverwriteSampleFromDatasetWeights(self):', 'pipeline_config', '=', 'pipeline_pb2.TrainEvalPipelineConfig()', 'pipeline_config.train_input_reader.sample_from_datasets_weights.extend([1,', '2])', 'pipeline_config_path', '=', 'os.path.join(self.get_temp_dir(),', "'pipeline.config')", '_write_config(pipeline_c...
975,401