Quick Image Classifier #1: AI Learns to Recognize Doodles with Machine Learning

Episode 1 of the video tutorial series on making an image classifier for doodle recognition. Learn more about machine learning algorithms used in this project on my blog: https://www.askforgametask.com/tutorial/machine-learning/image-classifier-doodle-recognition-cnn-intro/ 00:00 - Intro 00:14 - Playing "Quick, Draw!" game 00:44 - Project Specifications 01:28 - Playing Doodle Predictor game 02:37 - End Screen ---------------------------------------------------------------------------------------------------------- ★ Tutorial Info ★ In this 4-episode tutorial series, I'm going to show you how to do image classification using TensorFlow to make a game like Quick Draw! As you know, Quick Draw! is an online game that challenges players to draw a doodle and then artificial intelligence guesses what the drawings represent. Using Phaser 2 framework (http://phaser.io/) for making HTML5 games and TensorFlow JS (https://js.tensorflow.org) for implementing machine learning in Javascript, I made a similar game called Doodle Predictor. It's actually an image classifier app that runs directly in the browser. To perform image recognition, the game uses Artificial Intelligence (AI) powered by Machine Learning (ML) and Convolutional Neural Network (CNN). To avoid loading big data files, the model is trained on a small subset of the Quick Draw Dataset using only 500 images per each category. This means that the output predictions will not always be correct because the model is not trained for each possible case. In order to get better and more accurate predictions from an image classifier, its convolution neural network should be trained using much more images. But for learning purposes this is quite enough. Play Doodle Predictor here: https://www.askforgametask.com/tutorial/machine-learning/image-classifier-doodle-recognition-cnn-intro/ Get the source code here: https://github.com/ssusnic/Machine-Learning-Doodle-Recognition ---------------------------------------------------------------------------------------------------------- ★ About Episode 1 ★ The first episode introduces us to the whole project. It starts with playing the original "Quick, Draw!" game. Then it shows the main steps of creating a game based on image classification using tensorflow. The episode ends with a demo of the complete Doodle Predictor game that recognizes three different doodle categories: fish, car and snowman. ---------------------------------------------------------------------------------------------------------- ★ Playlist ★ To watch the entire series, check out the full playlist: https://www.youtube.com/watch?v=kLF5vx5Ya1A&list=PLiUmDJi4BtJtnbc-rzGmz7odFMzbbeCYE Here is a list of all the episodes with the covered topics: ► Episode 1: https://youtu.be/kLF5vx5Ya1A • Introduction ► Episode 2: https://youtu.be/-iBBWDJrUEs • Part 1 - Project Setup • Part 2 - Getting Data • Part 3 - Building a Convolutional Neural Network Model • Part 4 - Training the Model ► Episode 3: https://youtu.be/rCogVEJe3sU • Part 5 - Predicting Samples • Part 6 - Drawing Doodles • Part 7 - Recognizing Doodles • Part 8 - Adding More Doodle Categories ► Episode 4: (coming soon) • Game Demo ---------------------------------------------------------------------------------------------------------- ★ Contact Info ★ Official Website: https://www.askforgametask.com Facebook: https://www.facebook.com/askforgametask Twitter: https://twitter.com/ssusnic Github: https://github.com/ssusnic/ ---------------------------------------------------------------------------------------------------------- ★ Credits ★ Background Music: 'Pixelland' by Kevin MacLeod incompetech.com Intro Music: 'Euro Pop 80s' by Jay Man www.ourmusicbox.com