Teachable Machine: Approachable Web-Based Tool for Exploring Machine Learning Classification

By Michelle Carney, Barron Webster, Irene Alvarado, K. Phillips, Noura Howell, Jordan Griffith, J. Jongejan, Amit Pitaru, Alexander Chen. CHI Extended Abstracts, 2020

331 citations recorded by the source corpus.

Teachable Machine (teachablemachine.withgoogle.com) is a web-based GUI tool for creating custom machine learning classification models without specialized technical expertise. (Machine learning, or ML, lets systems learn to analyze data without being explicitly programmed.) We created it to help students, teachers, designers, and others learn about ML by creating and using their own classification models. Its broad uptake suggests it has empowered people to learn, teach, and explore ML concepts: People have created curriculum, tutorials, and other resources using Teachable Machine on topics like AI ethics at institutions including the Stanford d.school, NYU's Interactive Telecommunications Program, the MIT Media Lab, as well as creative experiments. Users in 201 countries have created over 125,000 classification models. Here we outline the project and its key contributions of (1) a flexible, approachable interface for ML classification models without ML or coding expertise, (2) a set of technical and design decisions that can inform future interactive machine learning tools, and (3) an example of how structured learning content surrounding the tool supports people accessing ML concepts.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/4a17a8ab29ecffde5e8280382dd858c75d5bb6b6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Toward User-Driven Sound Recognizer Personalization with People Who Are d/Deaf or Hard of Hearing (2021, methodology): “After collecting the minimum samples required by Teachable Machine for each class—20 for background noise and eight each for hand claps and paper crumpling—we invited participants to share their interpretation of the spectrogram audio representations.1 The data was then passed to Teachable Machine’s”
CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilities (2022, background): “For example, Arduino made accessible the interaction design possibilities of sensors and motors [39]; Wekinator [22] and Teachable Machine [11] made accessible the otherwise abstract capabilities of supervised machine learning classifers.”
CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualization (2020, background): “Teachable Machine [9] teaches the basic concept of machine learning classification, such as overfitting and underfitting, by allowing users to train a deep neural network classifier with data collected from their own webcam or microphone.”
AI Transparency in the Age of LLMs: A Human-Centered Research Roadmap (2023, background): “This is well reflected in the interdisciplinary area of interactive machine learning (iML) [4]—learning interactively through feedback from end-users—and related areas such as machine teaching [31, 170].”