GitLab

250447559  by Zhichao Lu:

    Update expected files format for Instance Segmentation challenge:
    - add fields ImageWidth, ImageHeight and store the values per prediction
    - as mask, store only encoded image and assume its size is ImageWidth x ImageHeight

--
250402780  by rathodv:

    Fix failing Mask R-CNN TPU convergence test.

    Cast second stage prediction tensors from bfloat16 to float32 to prevent errors in third target assignment (Mask Prediction) - Concat with different types bfloat16 and bfloat32 isn't allowed.

--
250300240  by Zhichao Lu:

    Addion Open Images Challenge 2019 object detection and instance segmentation
    support into Estimator framework.

--
249944839  by rathodv:

    Modify exporter.py to add multiclass score nodes in exported inference graphs.

--
249935201  by rathodv:

    Modify postprocess methods to preserve multiclass scores after non max suppression.

--
249878079  by Zhichao Lu:

    This CL slightly refactors some Object Detection helper functions for data creation, evaluation, and groundtruth providing.

    This will allow the eager+function custom loops to share code with the existing estimator training loops.

    Concretely we make the following changes:
    1. In input creation we separate dataset-creation into top-level helpers, and allow it to optionally accept a pre-constructed model directly instead of always creating a model from the config just for feature preprocessing.

    2. In coco evaluation we split the update_op creation into its own function, which the custom loops will call directly.

    3. In model_lib we move groundtruth providing/ datastructure munging into a helper function

    4. For now we put an escape hatch in `_summarize_target_assignment` when executing in tf v2.0 behavior because the summary apis used only work w/ tf 1.x

--
249673507  by rathodv:

    Use explicit casts instead of tf.to_float and tf.to_int32 to avoid warnings.

--
249656006  by Zhichao Lu:

    Add named "raw_keypoint_locations" node that corresponds with the "raw_box_locations" node.

--
249651674  by rathodv:

    Keep proposal boxes in float format. MatMulCropAndResize can handle the type even when feature themselves are bfloat16s.

--
249568633  by rathodv:

    Support q > 1 in class agnostic NMS.
    Break post_processing_test.py into 3 separate files to avoid linter errors.

--
249535530  by rathodv:

    Update some deprecated arguments to tf ops.

--
249368223  by rathodv:

    Modify MatMulCropAndResize to use MultiLevelRoIAlign method and move the tests to spatial_transform_ops.py module.

    This cl establishes that CropAndResize and RoIAlign are equivalent and only differ in the sampling point grid within the boxes. CropAndResize uses a uniform size x size point grid such that the corner points exactly overlap box corners, while RoiAlign divides boxes into size x size cells and uses their centers as sampling points. In this cl, we switch MatMulCropAndResize to use the MultiLevelRoIAlign implementation with `align_corner` option as MultiLevelRoIAlign implementation is more memory efficient on TPU when compared to the original MatMulCropAndResize.

--
249337338  by chowdhery:

    Add class-agnostic non-max-suppression in post_processing

--
249139196  by Zhichao Lu:

    Fix positional argument bug in export_tflite_ssd_graph

--
249120219  by Zhichao Lu:

    Add evaluator for computing precision limited to a given recall range.

--
249030593  by Zhichao Lu:

    Evaluation util to run segmentation and detection challenge evaluation.

--
248554358  by Zhichao Lu:

    This change contains the auxiliary changes required for TF 2.0 style training with eager+functions+dist strat loops, but not the loops themselves.

    It includes:
    - Updates to shape usage to support both tensorshape v1 and tensorshape v2
    - A fix to FreezableBatchNorm to not override the `training` arg in call when `None` was passed to the constructor (Not an issue in the estimator loops but it was in the custom loops)
    - Puts some constants in init_scope so they work in eager + functions
    - Makes learning rate schedules return a callable in eager mode (required so they update when the global_step changes)
    - Makes DetectionModel a tf.module so it tracks variables (e.g. ones nested in layers)
    - Removes some references to `op.name` for some losses and replaces it w/ explicit names
    - A small part of the change to allow the coco evaluation metrics to work in eager mode

--
248271226  by rathodv:

    Add MultiLevel RoIAlign op.

--
248229103  by rathodv:

    Add functions to 1. pad features maps 2. ravel 5-D indices

--
248206769  by rathodv:

    Add utilities needed to introduce RoI Align op.

--
248177733  by pengchong:

    Internal changes

--
247742582  by Zhichao Lu:

    Open Images Challenge 2019 instance segmentation metric: part 2

--
247525401  by Zhichao Lu:

    Update comments on max_class_per_detection.

--
247520753  by rathodv:

    Add multilevel crop and resize operation that builds on top of matmul_crop_and_resize.

--
247391600  by Zhichao Lu:

    Open Images Challenge 2019 instance segmentation metric

--
247325813  by chowdhery:

    Quantized MobileNet v2 SSD FPNLite config with depth multiplier 0.75

--

PiperOrigin-RevId: 250447559

236813471  by lzc:

    Internal change.

--
236507310  by lzc:

    Fix preprocess.random_resize_method config type issue. The target height and width will be passed as "size" to tf.image.resize_images which only accepts integer.

--
236409989  by Zhichao Lu:

    Config export_to_tpu from function parameter instead of HParams for TPU inference.

--
236403186  by Zhichao Lu:

    Make graph file names optional arguments.

--
236237072  by Zhichao Lu:

    Minor bugfix for keyword args.

--
236209602  by Zhichao Lu:

    Add support for PartitionedVariable to get_variables_available_in_checkpoint.

--
235828658  by Zhichao Lu:

    Automatically stop evaluation jobs when training is finished.

--
235817964  by Zhichao Lu:

    Add an optional process_metrics_fn callback to eval_util, it gets called
    with evaluation results once each evaluation is complete.

--
235788721  by lzc:

    Fix yml file tf runtime version.

--
235262897  by Zhichao Lu:

    Add keypoint support to the random_pad_image preprocessor method.

--
235257380  by Zhichao Lu:

    Support InputDataFields.groundtruth_confidences in retain_groundtruth(), retain_groundtruth_with_positive_classes(), filter_groundtruth_with_crowd_boxes(), filter_groundtruth_with_nan_box_coordinates(), filter_unrecognized_classes().

--
235109188  by Zhichao Lu:

    Fix bug in pad_input_data_to_static_shapes for num_additional_channels > 0; make color-specific data augmentation only touch RGB channels.

--
235045010  by Zhichao Lu:

    Don't slice class_predictions_with_background when add_background_class is false.

--
235026189  by lzc:

    Fix import in g3doc.

--
234863426  by Zhichao Lu:

    Added fixes in exporter to allow writing a checkpoint to a specified temporary directory.

--
234671886  by lzc:

    Internal Change.

--
234630803  by rathodv:

    Internal Change.

--
233985896  by Zhichao Lu:

    Add Neumann optimizer to object detection.

--
233560911  by Zhichao Lu:

    Add NAS-FPN object detection with Resnet and Mobilenet v2.

--
233513536  by Zhichao Lu:

    Export TPU compatible object detection model

--
233495772  by lzc:

    Internal change.

--
233453557  by Zhichao Lu:

    Create Keras-based SSD+MobilenetV1 for object detection.

--
233220074  by lzc:

    Update release notes date.

--
233165761  by Zhichao Lu:

    Support depth_multiplier and min_depth in _SSDResnetV1FpnFeatureExtractor.

--
233160046  by lzc:

    Internal change.

--
232926599  by Zhichao Lu:

    [tf.data] Switching tf.data functions to use `defun`, providing an escape hatch to continue using the legacy `Defun`.

    There are subtle differences between the implementation of `defun` and `Defun` (such as resources handling or control flow) and it is possible that input pipelines that use control flow or resources in their functions might be affected by this change. To migrate majority of existing pipelines to the recommended way of creating functions in TF 2.0 world, while allowing (a small number of) existing pipelines to continue relying on the deprecated behavior, this CL provides an escape hatch.

    If your input pipeline is affected by this CL, it should apply the escape hatch by replacing `foo.map(...)` with `foo.map_with_legacy_function(...)`.

--
232891621  by Zhichao Lu:

    Modify faster_rcnn meta architecture to normalize raw detections.

--
232875817  by Zhichao Lu:

    Make calibration a post-processing step.

    Specifically:
    - Move the calibration config from pipeline.proto --> post_processing.proto
    - Edit post_processing_builder.py to return a calibration function. If no calibration config is provided, it None.
    - Edit SSD and FasterRCNN meta architectures to optionally call the calibration function on detection scores after score conversion and before NMS.

--
232704481  by Zhichao Lu:

    Edit calibration builder to build a function that will be used within a detection model's `postprocess` method, after score conversion and before non-maxima suppression.

    Specific Edits:
    - The returned function now accepts class_predictions_with_background as its argument instead of detection_scores and detection_classes.
    - Class-specific calibration was temporarily removed, as it requires more significant refactoring. Will be added later.

--
232615379  by Zhichao Lu:

    Internal change

--
232483345  by ronnyvotel:

    Making the use of bfloat16 restricted to TPUs.

--
232399572  by Zhichao Lu:

    Edit calibration builder and proto to support class-agnostic calibration.

    Specifically:
    - Edit calibration protos to include path to relevant label map if required for class-specific calibration. Previously, label maps were inferred from other parts of the pipeline proto; this allows all information required by the builder stay within the calibration proto and remove extraneous information from being passed with class-agnostic calibration.
    - Add class-agnostic protos to the calibration config.

    Note that the proto supports sigmoid and linear interpolation parameters, but the builder currently only supports linear interpolation.

--
231613048  by Zhichao Lu:

    Add calibration builder for applying calibration transformations from output of object detection models.

    Specifically:
    - Add calibration proto to support sigmoid and isotonic regression (stepwise function) calibration.
    - Add a builder to support calibration from isotonic regression outputs.

--
231519786  by lzc:

    model_builder test refactor.
    - removed proto text boilerplate in each test case and let them call a create_default_proto function instead.
    - consolidated all separate ssd model creation tests into one.
    - consolidated all separate faster rcnn model creation tests into one.
    - used parameterized test for testing mask rcnn models and use_matmul_crop_and_resize
    - added all failures test.

--
231448169  by Zhichao Lu:

    Return static shape as a constant tensor.

--
231423126  by lzc:

    Add a release note for OID v4 models.

--
231401941  by Zhichao Lu:

    Adding correct labelmap for the models trained on Open Images V4 (*oid_v4
    config suffix).

--
231320357  by Zhichao Lu:

    Add scope to Nearest Neighbor Resize op so that it stays in the same name scope as the original resize ops.

--
231257699  by Zhichao Lu:

    Switch to using preserve_aspect_ratio in tf.image.resize_images rather than using a custom implementation.

--
231247368  by rathodv:

    Internal change.

--
231004874  by lzc:

    Update documentations to use tf 1.12 for object detection API.

--
230999911  by rathodv:

    Use tf.batch_gather instead of ops.batch_gather

--
230999720  by huizhongc:

    Fix weight equalization test in ops_test.

--
230984728  by rathodv:

    Internal update.

--
230929019  by lzc:

    Add an option to replace preprocess operation with placeholder for ssd feature extractor.

--
230845266  by lzc:

    Require tensorflow version 1.12 for object detection API and rename keras_applications to keras_models

--
230392064  by lzc:

    Add RetinaNet 101 checkpoint trained on OID v4 to detection model zoo.

--
230014128  by derekjchow:

    This file was re-located below the tensorflow/lite/g3doc/convert

--
229941449  by lzc:

    Update SSD mobilenet v2 quantized model download path.

--
229843662  by lzc:

    Add an option to use native resize tf op in fpn top-down feature map generation.

--
229636034  by rathodv:

    Add deprecation notice to a few old parameters in train.proto

--
228959078  by derekjchow:

    Remove duplicate elif case in _check_and_convert_legacy_input_config_key

--
228749719  by rathodv:

    Minor refactoring to make exporter's `build_detection_graph` method public.

--
228573828  by rathodv:

    Mofity model.postprocess to return raw detections and raw scores.

    Modify, post-process methods in core/model.py and the meta architectures to export raw detection (without any non-max suppression) and raw multiclass score logits for those detections.

--
228420670  by Zhichao Lu:

    Add shims for custom architectures for object detection models.

--
228241692  by Zhichao Lu:

    Fix the comment on "losses_mask" in "Loss" class.

--
228223810  by Zhichao Lu:

    Support other_heads' predictions in WeightSharedConvolutionalBoxPredictor. Also remove a few unused parameters and fix a couple of comments in convolutional_box_predictor.py.

--
228200588  by Zhichao Lu:

    Add Expected Calibration Error and an evaluator that calculates the metric for object detections.

--
228167740  by lzc:

    Add option to use bounded activations in FPN top-down feature map generation.

--
227767700  by rathodv:

    Internal.

--
226295236  by Zhichao Lu:

    Add Open Image V4 Resnet101-FPN training config to third_party

--
226254842  by Zhichao Lu:

    Fix typo in documentation.

--
225833971  by Zhichao Lu:

    Option to have no resizer in object detection model.

--
225824890  by lzc:

    Fixes p3 compatibility for model_lib.py

--
225760897  by menglong:

    normalizer should be at least 1.

--
225559842  by menglong:

    Add extra logic filtering unrecognized classes.

--
225379421  by lzc:

    Add faster_rcnn_inception_resnet_v2_atrous_oid_v4 config to third_party

--
225368337  by Zhichao Lu:

    Add extra logic filtering unrecognized classes.

--
225341095  by Zhichao Lu:

    Adding Open Images V4 models to OD API model zoo and corresponding configs to the
    configs.

--
225218450  by menglong:

    Add extra logic filtering unrecognized classes.

--
225057591  by Zhichao Lu:

    Internal change.

--
224895417  by rathodv:

    Internal change.

--
224209282  by Zhichao Lu:

    Add two data augmentations to object detection: (1) Self-concat (2) Absolute pads.

--
224073762  by Zhichao Lu:

    Do not create tf.constant until _generate() is actually called in the object detector.

--

PiperOrigin-RevId: 236813471

Release iNaturalist Species-trained models, refactor of evaluation, box predictor for object detection. (#5289)

* Merged commit includes the following changes:
212389173  by Zhichao Lu:

    1. Replace tf.boolean_mask with tf.where

--
212282646  by Zhichao Lu:

    1. Fix a typo in model_builder.py and add a test to cover it.

--
212142989  by Zhichao Lu:

    Only resize masks in meta architecture if it has not already been resized in the input pipeline.

--
212136935  by Zhichao Lu:

    Choose matmul or native crop_and_resize in the model builder instead of faster r-cnn meta architecture.

--
211907984  by Zhichao Lu:

    Make eval input reader repeated field and update config util to handle this field.

--
211858098  by Zhichao Lu:

    Change the implementation of merge_boxes_with_multiple_labels.

--
211843915  by Zhichao Lu:

    Add Mobilenet v2 + FPN support.

--
211655076  by Zhichao Lu:

    Bug fix for generic keys in config overrides

    In generic configuration overrides, we had a duplicate entry for train_input_config and we were missing the eval_input_config and eval_config.

    This change also introduces testing for all config overrides.

--
211157501  by Zhichao Lu:

    Make the locally-modified conv defs a copy.

    So that it doesn't modify MobileNet conv defs globally for other code that
    transitively imports this package.

--
211112813  by Zhichao Lu:

    Refactoring visualization tools for Estimator's eval_metric_ops. This will make it easier for future models to take advantage of a single interface and mechanics.

--
211109571  by Zhichao Lu:

    A test decorator.

--
210747685  by Zhichao Lu:

    For FPN, when use_depthwise is set to true, use slightly modified mobilenet v1 config.

--
210723882  by Zhichao Lu:

    Integrating the losses mask into the meta architectures. When providing groundtruth, one can optionally specify annotation information (i.e. which images are labeled vs. unlabeled). For any image that is unlabeled, there is no loss accumulation.

--
210673675  by Zhichao Lu:

    Internal change.

--
210546590  by Zhichao Lu:

    Internal change.

--
210529752  by Zhichao Lu:

    Support batched inputs with ops.matmul_crop_and_resize.

    With this change the new inputs are images of shape [batch, heigh, width, depth] and boxes of shape [batch, num_boxes, 4]. The output tensor is of the shape [batch, num_boxes, crop_height, crop_width, depth].

--
210485912  by Zhichao Lu:

    Fix TensorFlow version check in object_detection_tutorial.ipynb

--
210484076  by Zhichao Lu:

    Reduce TPU memory required for single image matmul_crop_and_resize.

    Using tf.einsum eliminates intermediate tensors, tiling and expansion. for an image of size [40, 40, 1024] and boxes of shape [300, 4] HBM memory usage goes down from 3.52G to 1.67G.

--
210468361  by Zhichao Lu:

    Remove PositiveAnchorLossCDF/NegativeAnchorLossCDF to resolve "Main thread is not in main loop error" issue in local training.

--
210100253  by Zhichao Lu:

    Pooling pyramid feature maps: add option to replace max pool with convolution layers.

--
209995842  by Zhichao Lu:

    Fix a bug which prevents variable sharing in Faster RCNN.

--
209965526  by Zhichao Lu:

    Add support for enabling export_to_tpu through the estimator.

--
209946440  by Zhichao Lu:

    Replace deprecated tf.train.Supervisor with tf.train.MonitoredSession. MonitoredSession also takes away the hassle of starting queue runners.

--
209888003  by Zhichao Lu:

    Implement function to handle data where source_id is not set.

    If the field source_id is found to be the empty string for any image during runtime, it will be replaced with a random string. This avoids hash-collisions on dataset where many examples do not have source_id set. Those hash-collisions have unintended site effects and may lead to bugs in the detection pipeline.

--
209842134  by Zhichao Lu:

    Converting loss mask into multiplier, rather than using it as a boolean mask (which changes tensor shape). This is necessary, since other utilities (e.g. hard example miner) require a loss matrix with the same dimensions as the original prediction tensor.

--
209768066  by Zhichao Lu:

    Adding ability to remove loss computation from specific images in a batch, via an optional boolean mask.

--
209722556  by Zhichao Lu:

    Remove dead code.

    (_USE_C_API was flipped to True by default in TensorFlow 1.8)

--
209701861  by Zhichao Lu:

    This CL cleans-up some tf.Example creation snippets, by reusing the convenient tf.train.Feature building functions in dataset_util.

--
209697893  by Zhichao Lu:

    Do not overwrite num_epoch for eval input. This leads to errors in some cases.

--
209694652  by Zhichao Lu:

    Sample boxes by jittering around the currently given boxes.

--
209550300  by Zhichao Lu:

    `create_category_index_from_labelmap()` function now accepts `use_display_name` parameter.
    Also added create_categories_from_labelmap function for convenience

--
209490273  by Zhichao Lu:

    Check result_dict type before accessing image_id via key.

--
209442529  by Zhichao Lu:

    Introducing the capability to sample examples for evaluation. This makes it easy to specify one full epoch of evaluation, or a subset (e.g. sample 1 of every N examples).

--
208941150  by Zhichao Lu:

    Adding the capability of exporting the results in json format.

--
208888798  by Zhichao Lu:

    Fixes wrong dictionary key for num_det_boxes_per_image.

--
208873549  by Zhichao Lu:

    Reduce the number of HLO ops created by matmul_crop_and_resize.

    Do not unroll along the channels dimension. Instead, transpose the input image dimensions, apply tf.matmul and transpose back.

    The number of HLO instructions for 1024 channels reduce from 12368 to 110.

--
208844315  by Zhichao Lu:

    Add an option to use tf.non_maximal_supression_padded in SSD post-process

--
208731380  by Zhichao Lu:

    Add field in box_predictor config to enable mask prediction and update builders accordingly.

--
208699405  by Zhichao Lu:

    This CL creates a keras-based multi-resolution feature map extractor.

--
208557208  by Zhichao Lu:

    Add TPU tests for Faster R-CNN Meta arch.

    * Tests that two_stage_predict and total_loss tests run successfully on TPU.
    * Small mods to multiclass_non_max_suppression to preserve static shapes.

--
208499278  by Zhichao Lu:

    This CL makes sure the Keras convolutional box predictor & head layers apply activation layers *after* normalization (as opposed to before).

--
208391694  by Zhichao Lu:

    Updating visualization tool to produce multiple evaluation images.

--
208275961  by Zhichao Lu:

    This CL adds a Keras version of the Convolutional Box Predictor, as well as more general infrastructure for making Keras Prediction heads & Keras box predictors.

--
208275585  by Zhichao Lu:

    This CL enables the Keras layer hyperparameter object to build a dedicated activation layer, and to disable activation by default in the op layer construction kwargs.

    This is necessary because in most cases the normalization layer must be applied before the activation layer. So, in Keras models we must set the convolution activation in a dedicated layer after normalization is applied, rather than setting it in the convolution layer construction args.

--
208263792  by Zhichao Lu:

    Add a new SSD mask meta arch that can predict masks for SSD models.
    Changes including:
     - overwrite loss function to add mask loss computation.
     - update ssd_meta_arch to handle masks if predicted in predict and postprocessing.

--
208000218  by Zhichao Lu:

    Make FasterRCNN choose static shape operations only in training mode.

--
207997797  by Zhichao Lu:

    Add static boolean_mask op to box_list_ops.py and use that in faster_rcnn_meta_arch.py to support use_static_shapes option.

--
207993460  by Zhichao Lu:

    Include FGVC detection models in model zoo.

--
207971213  by Zhichao Lu:

    remove the restriction to run tf.nn.top_k op on CPU

--
207961187  by Zhichao Lu:

    Build the first stage NMS function in the model builder and pass it to FasterRCNN meta arch.

--
207960608  by Zhichao Lu:

    Internal Change.

--
207927015  by Zhichao Lu:

    Have an option to use the TPU compatible NMS op cl/206673787, in the batch_multiclass_non_max_suppression function. On setting pad_to_max_output_size to true, the output nmsed boxes are padded to be of length max_size_per_class.

    This can be used in first stage Region Proposal Network in FasterRCNN model by setting the first_stage_nms_pad_to_max_proposals field to true in config proto.

--
207809668  by Zhichao Lu:

    Add option to use depthwise separable conv instead of conv2d in FPN and WeightSharedBoxPredictor. More specifically, there are two related configs:
    - SsdFeatureExtractor.use_depthwise
    - WeightSharedConvolutionalBoxPredictor.use_depthwise

--
207808651  by Zhichao Lu:

    Fix the static balanced positive negative sampler's TPU tests

--
207798658  by Zhichao Lu:

    Fixes a post-refactoring bug where the pre-prediction convolution layers in the convolutional box predictor are ignored.

--
207796470  by Zhichao Lu:

    Make slim endpoints visible in FasterRCNNMetaArch.

--
207787053  by Zhichao Lu:

    Refactor ssd_meta_arch so that the target assigner instance is passed into the SSDMetaArch constructor rather than constructed inside.

--

PiperOrigin-RevId: 212389173

* Fix detection model zoo typo.

* Modify tf example decoder to handle label maps with either `display_name` or `name` fields seamlessly.

Currently, tf example decoder uses only `name` field to look up ids for class text field present in the data. This change uses both `display_name` and `name` fields in the label map to fetch ids for class text.

PiperOrigin-RevId: 212672223

* Modify create_coco_tf_record tool to write out class text instead of class labels.

PiperOrigin-RevId: 212679112

* Fix detection model zoo typo.

PiperOrigin-RevId: 212715692

* Adding the following two optional flags to WeightSharedConvolutionalBoxHead:
1) In the box head, apply clipping to box encodings in the box head.
2) In the class head, apply sigmoid to class predictions at inference time.

PiperOrigin-RevId: 212723242

* Support class confidences in merge boxes with multiple labels.

PiperOrigin-RevId: 212884998

* Creates multiple eval specs for object detection.

PiperOrigin-RevId: 212894556

* Set batch_norm on last layer in Mask Head to None.

PiperOrigin-RevId: 213030087

* Enable bfloat16 training for object detection models.

PiperOrigin-RevId: 213053547

* Skip padding op when unnecessary.

PiperOrigin-RevId: 213065869

* Modify `Matchers` to use groundtruth weights before performing matching.

Groundtruth weights tensor is used to indicate padding in groundtruth box tensor. It is handled in `TargetAssigner` by creating appropriate classification and regression target weights based on the groundtruth box each anchor matches to. However, options such as `force_match_all_rows` in `ArgmaxMatcher` force certain anchors to match to groundtruth boxes that are just paddings thereby reducing the number of anchors that could otherwise match to real groundtruth boxes.

For single stage models like SSD the effect of this is negligible as there are two orders of magnitude more anchors than the number of padded groundtruth boxes. But for Faster R-CNN and Mask R-CNN where there are only 300 anchors in the second stage, a significant number of these match to groundtruth paddings reducing the number of anchors regressing to real groundtruth boxes degrading the performance severely.

Therefore, this change introduces an additional boolean argument `valid_rows` to `Matcher.match` methods and the implementations now ignore such padded groudtruth boxes during matching.

PiperOrigin-RevId: 213345395

* Add release note for iNaturalist Species trained models.

PiperOrigin-RevId: 213347179

* Fix the bug of uninitialized gt_is_crowd_list variable.

PiperOrigin-RevId: 213364858

* ...text exposed to open source public git repo...

PiperOrigin-RevId: 213554260

For details see our paper:
"Speed/accuracy trade-offs for modern convolutional object detectors."
Huang J, Rathod V, Sun C, Zhu M, Korattikara A, Fathi A, Fischer I,
Wojna Z, Song Y, Guadarrama S, Murphy K, CVPR 2017
https://arxiv.org/abs/1611.10012