Learning and the Unknown: Surveying Steps toward Open World Recognition


  • T. E. Boult University of Colorado
  • S. Cruz University of Colorado
  • A.R. Dhamija University of Colorado
  • M. Gunther University of Colorado
  • J. Henrydoss University of Colorado
  • W.J. Scheirer University of Notre Dame




As science attempts to close the gap between man and machine by building systems capable of learning, we must embrace the importance of the unknown. The ability to differentiate between known and unknown can be considered a critical element of any intelligent self-learning system. The ability to reject uncertain inputs has a very long history in machine learning, as does including a background or garbage class to account for inputs that are not of interest. This paper explains why neither of these is genuinely sufficient for handling unknown inputs – uncertain is not unknown, and unknowns need not appear to be uncertain to a learning system. The past decade has seen the formalization and development of many open set algorithms, which provably bound the risk from unknown classes. We summarize the state of the art, core ideas, and results and explain why, despite the efforts to date, the current techniques are genuinely insufficient for handling unknown inputs, especially for deep networks.




How to Cite

Boult, T. E., Cruz, S., Dhamija, A., Gunther, M., Henrydoss, J., & Scheirer, W. (2019). Learning and the Unknown: Surveying Steps toward Open World Recognition. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01), 9801-9807. https://doi.org/10.1609/aaai.v33i01.33019801



Senior Member Presentation Track: Summary Talks