models/research/object_detection/README.md

![TensorFlow Requirement: 1.15](https://img.shields.io/badge/TensorFlow%20Requirement-1.15-brightgreen)
![TensorFlow 2 Not Supported](https://img.shields.io/badge/TensorFlow%202%20Not%20Supported-%E2%9C%95-red.svg)

# Tensorflow Object Detection API

Creating accurate machine learning models capable of localizing and identifying
multiple objects in a single image remains a core challenge in computer vision.
The TensorFlow Object Detection API is an open source framework built on top of
TensorFlow that makes it easy to construct, train and deploy object detection
models. At Google we’ve certainly found this codebase to be useful for our
computer vision needs, and we hope that you will as well. <p align="center">
<img src="g3doc/img/kites_detections_output.jpg" width=676 height=450> </p>
Contributions to the codebase are welcome and we would love to hear back from
you if you find this API useful. Finally if you use the Tensorflow Object
Detection API for a research publication, please consider citing:

```
"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
```

\[[link](https://arxiv.org/abs/1611.10012)\]\[[bibtex](https://scholar.googleusercontent.com/scholar.bib?q=info:l291WsrB-hQJ:scholar.google.com/&output=citation&scisig=AAGBfm0AAAAAWUIIlnPZ_L9jxvPwcC49kDlELtaeIyU-&scisf=4&ct=citation&cd=-1&hl=en&scfhb=1)\]

<p align="center">
  <img src="g3doc/img/tf-od-api-logo.png" width=140 height=195>
</p>

## Maintainers

Name           | GitHub
-------------- | ---------------------------------------------
Jonathan Huang | [jch1](https://github.com/jch1)
Vivek Rathod   | [tombstone](https://github.com/tombstone)
Ronny Votel    | [ronnyvotel](https://github.com/ronnyvotel)
Derek Chow     | [derekjchow](https://github.com/derekjchow)
Chen Sun       | [jesu9](https://github.com/jesu9)
Menglong Zhu   | [dreamdragon](https://github.com/dreamdragon)
Alireza Fathi  | [afathi3](https://github.com/afathi3)
Zhichao Lu     | [pkulzc](https://github.com/pkulzc)

## Table of contents

Setup:

*   <a href='g3doc/installation.md'>Installation</a><br>

Quick Start:

*   <a href='object_detection_tutorial.ipynb'>
      Quick Start: Jupyter notebook for off-the-shelf inference</a><br>
*   <a href="g3doc/running_pets.md">Quick Start: Training a pet detector</a><br>

Customizing a Pipeline:

*   <a href='g3doc/configuring_jobs.md'>
      Configuring an object detection pipeline</a><br>
*   <a href='g3doc/preparing_inputs.md'>Preparing inputs</a><br>

Running:

*   <a href='g3doc/running_locally.md'>Running locally</a><br>
*   <a href='g3doc/running_on_cloud.md'>Running on the cloud</a><br>

Extras:

*   <a href='g3doc/detection_model_zoo.md'>Tensorflow detection model zoo</a><br>
*   <a href='g3doc/exporting_models.md'>
      Exporting a trained model for inference</a><br>
*   <a href='g3doc/tpu_exporters.md'>
      Exporting a trained model for TPU inference</a><br>
*   <a href='g3doc/defining_your_own_model.md'>
      Defining your own model architecture</a><br>
*   <a href='g3doc/using_your_own_dataset.md'>
      Bringing in your own dataset</a><br>
*   <a href='g3doc/evaluation_protocols.md'>
      Supported object detection evaluation protocols</a><br>
*   <a href='g3doc/oid_inference_and_evaluation.md'>
      Inference and evaluation on the Open Images dataset</a><br>
*   <a href='g3doc/instance_segmentation.md'>
      Run an instance segmentation model</a><br>
*   <a href='g3doc/challenge_evaluation.md'>
      Run the evaluation for the Open Images Challenge 2018/2019</a><br>
*   <a href='g3doc/tpu_compatibility.md'>
      TPU compatible detection pipelines</a><br>
*   <a href='g3doc/running_on_mobile_tensorflowlite.md'>
      Running object detection on mobile devices with TensorFlow Lite</a><br>
*   <a href='g3doc/context_rcnn.md'>
      Context R-CNN documentation for data preparation, training, and export</a><br>

## Getting Help

To get help with issues you may encounter using the Tensorflow Object Detection
API, create a new question on [StackOverflow](https://stackoverflow.com/) with
the tags "tensorflow" and "object-detection".

Please report bugs (actually broken code, not usage questions) to the
tensorflow/models GitHub
[issue tracker](https://github.com/tensorflow/models/issues), prefixing the
issue name with "object_detection".

Please check [FAQ](g3doc/faq.md) for frequently asked questions before reporting
an issue.

## Release information
### June 17th, 2020

We have released [Context R-CNN](https://arxiv.org/abs/1912.03538), a model that
uses attention to incorporate contextual information images (e.g. from
temporally nearby frames taken by a static camera) in order to improve accuracy.
Importantly, these contextual images need not be labeled.

*   When applied to a challenging wildlife detection dataset ([Snapshot Serengeti](http://lila.science/datasets/snapshot-serengeti)),
    Context R-CNN with context from up to a month of images outperforms a
    single-frame baseline by 17.9% mAP, and outperforms S3D (a 3d convolution
    based baseline) by 11.2% mAP.
*   Context R-CNN leverages temporal context from the unlabeled frames of a
    novel camera deployment to improve performance at that camera, boosting
    model generalizeability.

We have provided code for generating data with associated context
[here](g3doc/context_rcnn.md), and a sample config for a Context R-CNN
model [here](samples/configs/context_rcnn_resnet101_snapshot_serengeti_sync.config).

Snapshot Serengeti-trained Faster R-CNN and Context R-CNN models can be found in
the [model zoo](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md#snapshot-serengeti-camera-trap-trained-models).

A colab demonstrating Context R-CNN is provided
[here](colab_tutorials/context_rcnn_tutorial.ipynb).

<b>Thanks to contributors</b>: Sara Beery, Jonathan Huang, Guanhang Wu, Vivek
Rathod, Ronny Votel, Zhichao Lu, David Ross, Pietro Perona, Tanya Birch, and
the Wildlife Insights AI Team.

### May 19th, 2020

We have released [MobileDets](https://arxiv.org/abs/2004.14525), a set of
high-performance models for mobile CPUs, DSPs and EdgeTPUs.

*   MobileDets outperform MobileNetV3+SSDLite by 1.7 mAP at comparable mobile
    CPU inference latencies. MobileDets also outperform MobileNetV2+SSDLite by
    1.9 mAP on mobile CPUs, 3.7 mAP on EdgeTPUs and 3.4 mAP on DSPs while
    running equally fast. MobileDets also offer up to 2x speedup over MnasFPN on
    EdgeTPUs and DSPs.

For each of the three hardware platforms we have released model definition,
model checkpoints trained on the COCO14 dataset and converted TFLite models in
fp32 and/or uint8.

<b>Thanks to contributors</b>: Yunyang Xiong, Hanxiao Liu, Suyog Gupta, Berkin
Akin, Gabriel Bender, Pieter-Jan Kindermans, Mingxing Tan, Vikas Singh, Bo Chen,
Quoc Le, Zhichao Lu.

### May 7th, 2020

We have released a mobile model with the
[MnasFPN head](https://arxiv.org/abs/1912.01106).

*   MnasFPN with MobileNet-V2 backbone is the most accurate (26.6 mAP at 183ms
    on Pixel 1) mobile detection model we have released to date. With
    depth-multiplier, MnasFPN with MobileNet-V2 backbone is 1.8 mAP higher than
    MobileNet-V3-Large with SSDLite (23.8 mAP vs 22.0 mAP) at similar latency
    (120ms) on Pixel 1.

We have released model definition, model checkpoints trained on the COCO14
dataset and a converted TFLite model.

<b>Thanks to contributors</b>: Bo Chen, Golnaz Ghiasi, Hanxiao Liu, Tsung-Yi
Lin, Dmitry Kalenichenko, Hartwig Adam, Quoc Le, Zhichao Lu, Jonathan Huang, Hao
Xu.

### Nov 13th, 2019

We have released MobileNetEdgeTPU SSDLite model.

*   SSDLite with MobileNetEdgeTPU backbone, which achieves 10% mAP higher than
    MobileNetV2 SSDLite (24.3 mAP vs 22 mAP) on a Google Pixel4 at comparable
    latency (6.6ms vs 6.8ms).

Along with the model definition, we are also releasing model checkpoints trained
on the COCO dataset.

<b>Thanks to contributors</b>: Yunyang Xiong, Bo Chen, Suyog Gupta, Hanxiao Liu,
Gabriel Bender, Mingxing Tan, Berkin Akin, Zhichao Lu, Quoc Le

### Oct 15th, 2019

We have released two MobileNet V3 SSDLite models (presented in
[Searching for MobileNetV3](https://arxiv.org/abs/1905.02244)).

*   SSDLite with MobileNet-V3-Large backbone, which is 27% faster than Mobilenet
    V2 SSDLite (119ms vs 162ms) on a Google Pixel phone CPU at the same mAP.
*   SSDLite with MobileNet-V3-Small backbone, which is 37% faster than MnasNet
    SSDLite reduced with depth-multiplier (43ms vs 68ms) at the same mAP.

Along with the model definition, we are also releasing model checkpoints trained
on the COCO dataset.

<b>Thanks to contributors</b>: Bo Chen, Zhichao Lu, Vivek Rathod, Jonathan Huang

### July 1st, 2019

We have released an updated set of utils and an updated
[tutorial](g3doc/challenge_evaluation.md) for all three tracks of the
[Open Images Challenge 2019](https://storage.googleapis.com/openimages/web/challenge2019.html)!

The Instance Segmentation metric for
[Open Images V5](https://storage.googleapis.com/openimages/web/index.html) and
[Challenge 2019](https://storage.googleapis.com/openimages/web/challenge2019.html)
is part of this release. Check out
[the metric description](https://storage.googleapis.com/openimages/web/evaluation.html#instance_segmentation_eval)
on the Open Images website.

<b>Thanks to contributors</b>: Alina Kuznetsova, Rodrigo Benenson

### Feb 11, 2019

We have released detection models trained on the Open Images Dataset V4 in our
detection model zoo, including

*   Faster R-CNN detector with Inception Resnet V2 feature extractor
*   SSD detector with MobileNet V2 feature extractor
*   SSD detector with ResNet 101 FPN feature extractor (aka RetinaNet-101)

<b>Thanks to contributors</b>: Alina Kuznetsova, Yinxiao Li

### Sep 17, 2018

We have released Faster R-CNN detectors with ResNet-50 / ResNet-101 feature
extractors trained on the
[iNaturalist Species Detection Dataset](https://github.com/visipedia/inat_comp/blob/master/2017/README.md#bounding-boxes).
The models are trained on the training split of the iNaturalist data for 4M
iterations, they achieve 55% and 58% mean AP@.5 over 2854 classes respectively.
For more details please refer to this [paper](https://arxiv.org/abs/1707.06642).

<b>Thanks to contributors</b>: Chen Sun

### July 13, 2018

There are many new updates in this release, extending the functionality and
capability of the API:

*   Moving from slim-based training to
    [Estimator](https://www.tensorflow.org/api_docs/python/tf/estimator/Estimator)-based
    training.
*   Support for [RetinaNet](https://arxiv.org/abs/1708.02002), and a
    [MobileNet](https://ai.googleblog.com/2017/06/mobilenets-open-source-models-for.html)
    adaptation of RetinaNet.
*   A novel SSD-based architecture called the
    [Pooling Pyramid Network](https://arxiv.org/abs/1807.03284) (PPN).
*   Releasing several [TPU](https://cloud.google.com/tpu/)-compatible models.
    These can be found in the `samples/configs/` directory with a comment in the
    pipeline configuration files indicating TPU compatibility.
*   Support for quantized training.
*   Updated documentation for new binaries, Cloud training, and
    [Tensorflow Lite](https://www.tensorflow.org/mobile/tflite/).

See also our
[expanded announcement blogpost](https://ai.googleblog.com/2018/07/accelerated-training-and-inference-with.html)
and accompanying tutorial at the
[TensorFlow blog](https://medium.com/tensorflow/training-and-serving-a-realtime-mobile-object-detector-in-30-minutes-with-cloud-tpus-b78971cf1193).

<b>Thanks to contributors</b>: Sara Robinson, Aakanksha Chowdhery, Derek Chow,
Pengchong Jin, Jonathan Huang, Vivek Rathod, Zhichao Lu, Ronny Votel

### June 25, 2018

Additional evaluation tools for the
[Open Images Challenge 2018](https://storage.googleapis.com/openimages/web/challenge.html)
are out. Check out our short tutorial on data preparation and running evaluation
[here](g3doc/challenge_evaluation.md)!

<b>Thanks to contributors</b>: Alina Kuznetsova

### June 5, 2018

We have released the implementation of evaluation metrics for both tracks of the
[Open Images Challenge 2018](https://storage.googleapis.com/openimages/web/challenge.html)
as a part of the Object Detection API - see the
[evaluation protocols](g3doc/evaluation_protocols.md) for more details.
Additionally, we have released a tool for hierarchical labels expansion for the
Open Images Challenge: check out
[oid_hierarchical_labels_expansion.py](dataset_tools/oid_hierarchical_labels_expansion.py).

<b>Thanks to contributors</b>: Alina Kuznetsova, Vittorio Ferrari, Jasper
Uijlings

### April 30, 2018

We have released a Faster R-CNN detector with ResNet-101 feature extractor
trained on [AVA](https://research.google.com/ava/) v2.1. Compared with other
commonly used object detectors, it changes the action classification loss
function to per-class Sigmoid loss to handle boxes with multiple labels. The
model is trained on the training split of AVA v2.1 for 1.5M iterations, it
achieves mean AP of 11.25% over 60 classes on the validation split of AVA v2.1.
For more details please refer to this [paper](https://arxiv.org/abs/1705.08421).

<b>Thanks to contributors</b>: Chen Sun, David Ross

### April 2, 2018

Supercharge your mobile phones with the next generation mobile object detector!
We are adding support for MobileNet V2 with SSDLite presented in
[MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381).
This model is 35% faster than Mobilenet V1 SSD on a Google Pixel phone CPU
(200ms vs. 270ms) at the same accuracy. Along with the model definition, we are
also releasing a model checkpoint trained on the COCO dataset.

<b>Thanks to contributors</b>: Menglong Zhu, Mark Sandler, Zhichao Lu, Vivek
Rathod, Jonathan Huang

### February 9, 2018

We now support instance segmentation!! In this API update we support a number of
instance segmentation models similar to those discussed in the
[Mask R-CNN paper](https://arxiv.org/abs/1703.06870). For further details refer
to [our slides](http://presentations.cocodataset.org/Places17-GMRI.pdf) from the
2017 Coco + Places Workshop. Refer to the section on
[Running an Instance Segmentation Model](g3doc/instance_segmentation.md) for
instructions on how to configure a model that predicts masks in addition to
object bounding boxes.

<b>Thanks to contributors</b>: Alireza Fathi, Zhichao Lu, Vivek Rathod, Ronny
Votel, Jonathan Huang

### November 17, 2017

As a part of the Open Images V3 release we have released:

*   An implementation of the Open Images evaluation metric and the
    [protocol](g3doc/evaluation_protocols.md#open-images).
*   Additional tools to separate inference of detection and evaluation (see
    [this tutorial](g3doc/oid_inference_and_evaluation.md)).
*   A new detection model trained on the Open Images V2 data release (see
    [Open Images model](g3doc/detection_model_zoo.md#open-images-models)).

See more information on the
[Open Images website](https://github.com/openimages/dataset)!

<b>Thanks to contributors</b>: Stefan Popov, Alina Kuznetsova

### November 6, 2017

We have re-released faster versions of our (pre-trained) models in the
<a href='g3doc/detection_model_zoo.md'>model zoo</a>. In addition to what was
available before, we are also adding Faster R-CNN models trained on COCO with
Inception V2 and Resnet-50 feature extractors, as well as a Faster R-CNN with
Resnet-101 model trained on the KITTI dataset.

<b>Thanks to contributors</b>: Jonathan Huang, Vivek Rathod, Derek Chow, Tal
Remez, Chen Sun.

### October 31, 2017

We have released a new state-of-the-art model for object detection using the
Faster-RCNN with the
[NASNet-A image featurization](https://arxiv.org/abs/1707.07012). This model
achieves mAP of 43.1% on the test-dev validation dataset for COCO, improving on
the best available model in the zoo by 6% in terms of absolute mAP.

<b>Thanks to contributors</b>: Barret Zoph, Vijay Vasudevan, Jonathon Shlens,
Quoc Le

### August 11, 2017

We have released an update to the
[Android Detect demo](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/examples/android)
which will now run models trained using the Tensorflow Object Detection API on
an Android device. By default, it currently runs a frozen SSD w/Mobilenet
detector trained on COCO, but we encourage you to try out other detection
models!

<b>Thanks to contributors</b>: Jonathan Huang, Andrew Harp

### June 15, 2017

In addition to our base Tensorflow detection model definitions, this release
includes:

*   A selection of trainable detection models, including:
    *   Single Shot Multibox Detector (SSD) with MobileNet,
    *   SSD with Inception V2,
    *   Region-Based Fully Convolutional Networks (R-FCN) with Resnet 101,
    *   Faster RCNN with Resnet 101,
    *   Faster RCNN with Inception Resnet v2
*   Frozen weights (trained on the COCO dataset) for each of the above models to
    be used for out-of-the-box inference purposes.
*   A [Jupyter notebook](colab_tutorials/object_detection_tutorial.ipynb) for
    performing out-of-the-box inference with one of our released models
*   Convenient [local training](g3doc/running_locally.md) scripts as well as
    distributed training and evaluation pipelines via
    [Google Cloud](g3doc/running_on_cloud.md).

<b>Thanks to contributors</b>: Jonathan Huang, Vivek Rathod, Derek Chow, Chen
Sun, Menglong Zhu, Matthew Tang, Anoop Korattikara, Alireza Fathi, Ian Fischer,
Zbigniew Wojna, Yang Song, Sergio Guadarrama, Jasper Uijlings, Viacheslav
Kovalevskyi, Kevin Murphy