Source Code for Automate Generate tfrecord from Kaggle Grab and Go Dataset

set up your python environment and use package manager pip to install requirement depedency

pip install -r requirements.txt

then install tensorflow object detection api using our setupapi

git clone /~https://github.com/tensorflow/models
cd models/research/
protoc object_detection/protos/*.proto --python_out=.
cp object_detection/packages/tf2/setup.py .
python -m pip install .

Download the kaggle dataset from our collected retail data in this link

Extract the data so the project repository structure look like this

.
├── data/                   # CSV file location
├── annotations/            # Extracted Dataset
    └── train/
    └── test/
├── images/                 # Extracted Dataset
    └── train/
    └── test/                 
├── generate_tfrecord.py                     
├── requirements.txt                    
├── detection_label_map.pbtxt                   
├── LICENSE
└── README.md

to generate the tfrecords formated input we need to convert XML annotations data to CSV data by running

python xml_to_csv.py

We use tfrecords as input for our model training, to generate it we run generate_tfrecord.py by running

python generate_tfrecord.py --csv_input=data/train_labels.csv  --output_path=train.record --image_dir=images/train/
python generate_tfrecord.py --csv_input=data/test_labels.csv  --output_path=test.record --image_dir=images/test/

Access the Colab Notebook in this link to train Grab and Go Retail Detection Model