This is an implemtation of Multi Label Classification using Tensorflow. The model generates class names and scores about an image.
The repository includes :
- Source code of Multi Label Classificatio on MobileNet V2.
- Training code for Open Image Dataset V4.
- Jupyter notebook to visualize the recognition results.
- (train.py, mobilenet_v2.py, config.py, datapipe.py): These files contain the Multi Label Classification implementation.
*[train_class_name.csv](/data/train_class_name.csv]: This file contains the class names and class index for Open Image Dataset in train folder.
- Inspect_Model.ipynb This notebook provides visualizations of every step of the pipeline for recognizing multi label on image.
you need OID_V4(Open Image Dataset V4) and put the dataset in data/OIDv4, if you want to train this model. Run train.py