ai_image_classifier 1.0.0

A simple and powerful Flutter package for on-device image classification using TensorFlow Lite models. Supports custom models, image resizing, and normalization.

<<<<<<< HEAD

AI Image Classifier

A Flutter package for easy on-device image classification using TensorFlow Lite models.

Features

• Load TFLite models from assets.
• Classify images from bytes or file paths.
• Automatic image preprocessing (resizing and normalization).
• Works with common classification models like MobileNet, ResNet, etc.

Getting Started

1. Add Dependencies

Add ai_image_classifier to your pubspec.yaml:

dependencies:
  ai_image_classifier:
    git:
      url: https://github.com/swetakothari16/ai_image_classifier.git

2. Add Model and Labels

Place your .tflite model and labels.txt in your assets folder and register them:

flutter:
  assets:
    - assets/models/mobilenet_v1.tflite
    - assets/models/labels.txt

3. Usage

import 'package:ai_image_classifier/ai_image_classifier.dart';

final classifier = AiImageClassifier();

// Load model
await classifier.loadModel(
  modelPath: 'assets/models/mobilenet_v1.tflite',
  labelsPath: 'assets/models/labels.txt',
);

// Classify image from path
List<Classification> results = await classifier.classifyImagePath(imagePath);

// Print results
for (var res in results) {
  print("${res.label}: ${(res.confidence * 100).toStringAsFixed(1)}%");
}

// Dispose when done
classifier.dispose();

How to get a sample model?

You can download a pre-trained MobileNet model from TensorFlow Hub:

1. Go to TensorFlow Hub MobileNet.
2. Download the .tflite file.
3. Use the ImageNet labels provided in the example app's labels.txt.

License

MIT

ai_image_classifier

45b5afbb9e165e764d9172268b8f7b3279e80ff5