In this tutorial, we will show you how to build an object detector for a custom dataset. You can find the custom dataset in the input/images folder. The goal is to detect the square, circle, and triangle shapes in the images.

We adapted the Dockerfile provided by Tensorflow and prepared a docker container preloaded with all the necessary libraries to get you started quickly to build your own object detection using custom dataset. We modified the scripts provided by Tensorflow and from other excellent online tutorials eg. https://github.com/bourdakos1/Custom-Object-Detection to give an easy to follow step by step tutorial.

Install Docker:

https://docs.docker.com/v17.12/docker-for-mac/install/#install-and-run-docker-for-mac

Clone the repository (https://github.com/aiformankind/object-detection):

git clone https://github.com/aiformankind/object-detection.git

Go to the repository directory that you just clone:

cd object-detection

Build the Tensorflow docker (this job will pull the latest tensorflow images and set up the environment) :

docker build -t aiformankind/objectdetection:0.0.1 .

Start the Tensorflow container (this job will spin up the objectdetection container):

docker run -it -p 8888:8888 -p 6006:6006 --name=objectdetection aiformankind/objectdetection:0.0.1

Train model: You can stop the training by pressing CTRL-C after iteration step ~200.

python models/research/object_detection/legacy/train.py --logtostderr --train_dir=/tensorflow/train/ --pipeline_config_path=/tensorflow/train/faster_rcnn_resnet101.config

Create a frozen inference graph: Replace model.ckpt-XXX with the largest(latest) checkpoint file number. Look it up in the /tensorflow/train folder

python models/research/object_detection/export_inference_graph.py --input_type image_tensor --pipeline_config_path /tensorflow/train/faster_rcnn_resnet101.config --trained_checkpoint_prefix /tensorflow/train/model.ckpt-XXX --output_directory /tensorflow/inference_graph

Start jupyter notebook

jupyter notebook --ip 0.0.0.0 --no-browser --allow-root

You can access the jupyter notebook on browser. Find the object_detection_shapes.ipynb in /tensorflow/models/research/object-detection folder.

Run the notebook to see the our model in action. In the last cell, we use the model to detect and identify the square, circle, triangle in the test images.

First, you have to annotate your images by building bounding boxes around the objects you want to detect. There are many tools you can use. Among them are labelImg, RectLabel, and Labelbox.

You can find the custom images for this tutorial in the /tensorflow/input/images folder. The annotations of bounding boxes are described using XML format. These xmls are stored in /tensorflow/input/annotations/xmls folder. See XML element which describes the bounding box in the xml below. All the annotation tools mentioned above provide easy to use UI interface to draw the bounding boxes(annotations) and have the export option to create these XMLs automatically from your annotated images.

<annotation>
    <folder>tmpda5w5lkl</folder>
    <filename>cjov38btu6ya00975ow2d6t59.jpeg</filename>
    <path>/tmp/tmpda5w5lkl/cjov38btu6ya00975ow2d6t59.jpeg</path>
    <source>
        <database>Unknown</database>
    </source>
    <size>
        <width>1257</width>
        <height>1676</height>
        <depth>3</depth>
    </size>
    <segmented>0</segmented>
    <object>
        <name>Triangle</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
                <xmin>584</xmin>
                <ymin>449</ymin>
                <xmax>936</xmax>
                <ymax>794</ymax>
            </bndbox>
        </object>

1. mkdir modelBuild
2. cd modelBuild
3. curl -O "http://download.tensorflow.org/models/object_detection/faster_rcnn_resnet101_coco_11_06_2017.tar.gz"
4. tar xzf faster_rcnn_resnet101_coco_11_06_2017.tar.gz
5. python /notebooks/tensorflow/models/research/object_detection/legacy/train.py --logtostderr --train_dir=/notebooks/object-detection/modelBuild/build/ --pipeline_config_path=/notebooks/object-detection/faster_rcnn_resnet101.config
6. python /notebooks/tensorflow/models/research/object_detection/export_inference_graph.py --input_type image_tensor --pipeline_config_path /notebooks/object-detection/faster_rcnn_resnet101.config --trained_checkpoint_prefix /notebooks/object-detection/modelBuild/build/model.ckpt-XXX --output_directory /notebooks/object-detection/modelBuild/build/0