TensorFlow Lite Image Classifier

This project contains a Python script that utilizes a TensorFlow Lite model to classify images. The script loads a pre-trained TFLite model, processes an input image, and outputs the classification results.

Features

• Load a TensorFlow Lite model.
• Process images for model input.
• Classify images and output predictions.

Requirements

• Python 3.6 or newer
• TensorFlow
• NumPy
• Pillow (PIL)

Installation

To set up your environment to run this script, follow these steps:

1. Ensure Python 3.6+ is installed on your system.
2. Install the required Python packages:

pip install tensorflow numpy pillow

Usage

To use the script, you need to have a TFLite model file and a corresponding labels text file. Place your image file in the same directory as the script or specify the path to it.

Run the script using:

python main.py

Files Description

• main.py: Main script to load the model, process the image, and classify it.
• model_unquant1.tflite: TensorFlow Lite model file (ensure you have this file in the same directory).
• labels.txt: Text file containing labels corresponding to the model's output.
• perch.jpg: Example image file for testing the classifier.

Model Details

This script is configured to use a model named model_unquant1.tflite. Ensure that you replace "model_unquant1.tflite" and "labels.txt" in the script with the paths to your actual model and labels files, respectively.