🐶 Dog AI

demo.mp4

💡 Project

Recognize over 100 dog breeds by drag and drop an image using Tensorflow.js and Teachable Machine.

Warning

This project is solely built around Machine Learning (no LLMs, no servers), I trained the model myself with a public dataset, so please don't expect 100% accuracy.

🛠 Tools

• React
• React Dropzone
• Tensorflow.js
• Teachable Machine
• Vitest
• Biome
• React-Router v7

💻Demo

https://dogai.netlify.com

🚀Quick start

Installation

git clone git@github.com:jeferson-sb/dogAI.git && cd dogAI
npm install

Wikipedia integration

cp .env.example .env

VITE_WIKIPEDIA_ENDPOINT=https://en.wikipedia.org/w/api.php?origin=*&action=query&format=json&uselang=pt&prop=extracts&generator=prefixsearch&redirects=1&converttitles=1&formatversion=2&exintro=1&explaintext=1&gpssearch=
VITE_WIKIPEDIA_WIKI=https://en.wikipedia.org/wiki

Usage

npm run dev

Tests

npm run test

Linting

npm run lint
npm run format

How to Train your own model

1. Gather a dataset with a bunch of images
2. Resize and minify all the images
3. Separate dogs image by breed and rename all the files
4. Upload to Teachable Machine
5. Train your model
6. Export your trained model

Dataset Reference

Primary: Aditya Khosla, Nityananda Jayadevaprakash, Bangpeng Yao and Li Fei-Fei. Novel dataset for Fine-Grained Image Categorization. First Workshop on Fine-Grained Visual Categorization (FGVC), IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2011.

Secondary: J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li and L. Fei-Fei, ImageNet: A Large-Scale Hierarchical Image Database. IEEE Computer Vision and Pattern Recognition (CVPR), 2009.

📝License

This project is licensed under the MIT License