Building a Benchmark for Human-Level Concept Learning and Reasoning

Originally published at: https://developer.nvidia.com/blog/building-a-benchmark-for-human-level-concept-learning-and-reasoning/

Humans have an inherent ability to learn novel concepts from only a few samples and generalize these concepts to different situations. Even though today’s machine learning models excel with an abundance of training data on standard recognition tasks, a considerable gap exists between machine-level pattern recognition and human-level concept learning. Over 50 years ago, M.…

From early drawings of horses in the Lascaux Cave to the advanced Egyptian Hieroglyphic Alphabet - the ability for humankind to covey ideas through images and symbols persists across history. When will machine-level pattern recognition and human-level concept learning intersect? This latest benchmark works to close the gap. Share your thoughts on this work with the team.