Datasets:
celinelee
/
python_fn_calls

Dataset card Data Studio Files
xet
 Community
1
code
stringlengths
3
6.57k
tf.constant([proposal_boxes.shape[1]], dtype='int32')
box_list.BoxList(proposals_absolute)
get()
tf.stop_gradient(proposal_boxes)
tf.zeros((1, proposal_boxes.shape[1], 2)
tf.Print(proposal_boxes, [proposal_boxes], message="1: ")
self._format_groundtruth_data(true_image_shapes)
tf.Print(proposal_boxes, [proposal_boxes], message="2: ")
tf.Print(proposal_boxes, [], message=("Shape of pboxes " + str(proposal_boxes.shape[1])
tf.Print(num_proposals, [num_proposals])
tf.Print(proposal_boxes_normalized, [proposal_boxes_normalized], message="3: ")
tf.Print(proposal_boxes_normalized, [tf.shape(proposal_boxes_normalized)
tf.constant(self.proposals[:, 0:64, :], dtype='float32')
tf.Print(proposal_boxes_normalized, [], message=("Shape of minibatch " + str(proposal_boxes_normalized.shape[1])
tf.stop_gradient(flattened_proposal_feature_maps)
tf.Print(flattened_proposal_feature_maps, [], message=("Cropped props : " + str(flattened_proposal_feature_maps.shape)
_predict_third_stage(self, prediction_dict, image_shapes)
predicted (final)
predictions (logits)
predictions (at class index 0)
prediction_dict.update(detections_dict)
tf.shape(detection_boxes)
tf.shape(detection_boxes)
detection_masks.get_shape()
as_list()
detection_classes.get_shape()
as_list()
_gather_instance_masks(self, instance_masks, classes)
instance_masks.get_shape()
as_list()
tf.shape(instance_masks)
tf.reshape(instance_masks, [-1, height, width])
tf.to_int32(tf.reshape(classes, [-1])
tf.range(k)
tf.gather(instance_masks, gather_idx)
_extract_rpn_feature_maps(self, preprocessed_inputs)
anchors (for the RPN)
tf.shape(preprocessed_inputs)
tf.shape(rpn_features_to_crop)
slim.arg_scope(self._first_stage_box_predictor_arg_scope_fn()
_predict_rpn_proposals(self, rpn_box_predictor_features)
predictions (logits)
predictions (at class index 0)
self._first_stage_anchor_generator.num_anchors_per_location()
len(num_anchors_per_location)
return (tf.squeeze(box_encodings, axis=2)
that (partially)
predictions (logits)
predictions (at class index 0)
anchors (for the RPN)
predictions (logits)
predictions (at class index 0)
anchors (for the RPN)
_batch_gather_kept_indices(predictions_tensor)
partial(tf.gather, indices=keep_indices)
return (_batch_gather_kept_indices(box_encodings)
_batch_gather_kept_indices(objectness_predictions_with_background)
_flatten_first_two_dimensions(self, inputs)
shape_utils.combined_static_and_dynamic_shape(inputs)
tf.reshape(inputs, flattened_shape)
postprocess(self, prediction_dict, true_image_shapes)
remapped (and may thus have a different interpretation)
tf.name_scope('FirstStagePostprocessor')
tf.to_float(num_proposals)
TODO(jrru)
tf.name_scope('SecondStagePostprocessor')
or (self._number_of_stages == 3 and self._is_training)
prediction_dict.get(box_predictor.MASK_PREDICTIONS)
tf.sigmoid(prediction_dict[key])
predictions (logits)
tf.expand_dims(rpn_box_encodings_batch, axis=2)
tf.expand_dims(anchors, 0)
tf.squeeze(proposal_boxes, axis=2)
self._compute_clip_window(image_shapes)
tf.expand_dims(proposal_boxes, axis=2)
tf.stop_gradient(proposal_boxes)
self._format_groundtruth_data(true_image_shapes)
normalize_boxes(args)
box_list.BoxList(proposal_boxes_per_image)
get()
containing (absolute)
hot (or k-hot)
tf.unstack(proposal_boxes)
tf.unstack(proposal_scores)
tf.unstack(num_proposals)
box_list.BoxList(single_image_proposal_boxes)
sampled_boxlist.set(tf.Print(sampled_boxlist.get()
sampled_boxlist.num_boxes()
sampled_boxlist.num_boxes()
sampled_padded_boxlist.get()
tf.Print(bb, [single_image_groundtruth_boxlist.num_boxes()
str(bb.shape)
single_image_proposal_box_sample.append(bb)
sampled_padded_boxlist.get_field(fields.BoxListFields.scores)
return (tf.stack(single_image_proposal_box_sample)
tf.stack(single_image_proposal_score_sample)
tf.stack(single_image_num_proposals_sample)
_format_groundtruth_data(self, true_image_shapes, stage='detection')
containing (absolute)
hot (or k-hot)