Two years ago, Google launched a site called Teachable Machine, which let you train a simple model using their camera without any code. Now, it’s launching an updated version so you can train more advanced models. The earlier version allowed you to train three classes through your camera. The new model, not only lets you define more than three classes, it also allows you to use images, audio clips, pose data, or your own dataset for the training. Here are some of the most prominent features of the updated Teachable Machine:
Train a model on image data Train a model on audio data Train a model on pose data (like in Posenet) Upload your own datasets to train models Train more than 3 classes per model Disable classes Save your TensorFlow.js model Download your model Deploy your model to use in your own project (websites, apps, physical machines, etc.)
Google said some folks have already built some interesting projects using the Teachable Machine. For example, technology expert Steve Sailing built a communication tool for people with speech imparity. Meanwhile, Yining Shi, a professor at the Interactive Telecommunications Program at New York University, and her students, built a gaming model that was trained to move using different gestures.
Full project: https://t.co/07JK5LRVpZhttps://t.co/wZn7ePFCPp — Irene Alvarado (@ire_alva) May 22, 2019 The company said you don’t have to worry about your data as the whole process takes place locally on your machine. Plus, since the model is powered by Google’s popular open-source machine learning library Tensorflow, you can easily plug it into apps or websites. This is quite useful for novices like me who haven’t ever coded for machine learning applications. You can probably create a custom AI assistant or a sorting app if you know a bit of coding. And even if you don’t, this seems like a fun way to spend to learn about one of the hottest tech trends. You can check out the updated version of the Teachable Machine here.
