Gym Equipment Classification

This repository contains resources for a machine learning project focused on classifying gym equipment using TensorFlow. It includes datasets, pre-trained models, and Jupyter notebooks for building, training, fine-tuning, evaluating, and converting models into a format suitable for deployment on mobile devices.

Project Overview

The aim of this project is to classify 8 different types of gym equipment. The model is built using transfer learning with MobileNetV2, followed by fine-tuning to optimize performance for the specific task. The trained model is then saved in both .h5 (TensorFlow) and .tflite (TensorFlow Lite) formats to allow for deployment on different platforms.

Repository Structure

machine-learning-model
│
├── dataset/
│   ├── test/               # Test dataset
│   ├── train/              # Training dataset
│   ├── validation/         # Validation dataset
│   └── README.md           # Dataset documentation
│
├── model/
│   ├── gym_equip.classifier.tflite  # TensorFlow Lite model
│   ├── model_ft_finetuned.h5       # Fine-tuned TensorFlow model
│   └── model_metadata/
│       ├── gym_equip_classifier.json        # Model metadata in JSON format
│       └── gym_equip_classifier_metadata.tflite  # Metadata-enhanced TFLite model
│
├── notebooks/
│   ├── saved_model/         # Saved model directory
│   ├── evaluation.ipynb     # Model evaluation and testing (H5 and TFLite)
│   ├── exploratory.ipynb    # Exploratory Data Analysis (EDA) and visualization
│   └── modeling.ipynb       # Building, training, fine-tuning, saving, and converting TFLite model
│
├── scripts/
│   ├── README.md            # Scripts documentation
│   ├── labels.txt           # Labels for the classifier
│   └── metadata_writer_for_image_classifier.py  # Script to write metadata
│
├── README.md                # Project documentation
└── requirements.txt         # Required Python packages

Project Scope

• Classification Task: The model is designed to classify 8 types of gym equipment.
• Model Conversion: The trained model is converted into TensorFlow Lite (TFLite) format for mobile deployment.
• Transfer Learning: The project uses MobileNetV2 for transfer learning and applies fine-tuning techniques to improve accuracy on the gym equipment dataset.

Notebooks Overview

• exploratory.ipynb: Used for exploratory data analysis (EDA), data preprocessing, and visualization of the gym equipment dataset.
• modeling.ipynb: Contains the steps to build the model, train it using transfer learning (MobileNetV2), perform fine-tuning, and save the model. The model is then converted to TFLite format.
• Metadata Integration: Enhance the TFLite model with metadata for improved interpretability on edge devices.
• evaluation.ipynb: Used for evaluating and testing the model, both in its original .h5 (TensorFlow) and .tflite (TensorFlow Lite) formats.

Model Details

• model_ft_finetuned.h5: The final fine-tuned model saved in the TensorFlow .h5 format.
• gym_equip.classifier.tflite: The converted model in TensorFlow Lite format for edge device deployment.
• gym_equip_classifier_metadata.tflite: The TFLite model enhanced with metadata for easier use in applications.
• gym_equip_classifier.json: Metadata file describing the model's inputs, outputs, and labels.

Scripts Overview

• labels.txt: File containing the class labels for the gym equipment classifier.
• metadata_writer_for_image_classifier.py: Script to add metadata to the TFLite model.

Please refer to the scripts/README.md for more details on the scripts.

Dataset

The dataset is organized into three main folders:

• train/: Contains images used for training the model.
• test/: Contains images used for testing the model.
• validation/: Contains images used for validating the model during training.

Please refer to the dataset/README.md for more details on the dataset.

Installation

To get started, clone this repository and install the necessary dependencies:

1. Clone the repository

git clone https://github.com/dejavucapstone/machine-learning-model.git

cd machine-learning-model

Requirements

To run the code and experiments in this repository, you'll need the following Python packages:

# Install required packages
pip install -r requirements.txt