Round 7: Tossup 3

Belhumeur, Hespanha, and Krieger projected these things and then applied linear discriminant analysis to develop a method named for Ronald Fisher. Cascades of Haar features were used to identify these things in an early paper on boosting by Viola and Jones. Datasets are visualized using these things in a weird method proposed by Herman Chernoff. (15[1])An oft-taught (15[1])application of PCA produces blurry and punning “eigen” examples of these things. A company that hosts open-source (*) ML models and is known for its transformers library is named for a “Hugging” (10[2])one of these things. One of these things is the first object in the name of a company whose AI research group in 2016 developed PyTorch, and which is the major subsidiary of the creator of the LLM Llama. For 10 points, newer iPhones can use a camera to recognize what body parts for unlocking? ■END■

ANSWER: faces [accept facial recognition; accept Fisherface or eigenfaces or Hugging Face; accept Facebook; prompt on people or humans or equivalents; prompt on objects or images]
<TM, Other Science (Computer Science)> | NAFTA-Packet-7
= Average correct buzzpoint

Back to tossups

Buzzes


Summary

TournamentEditionMatchHeardConv. %Power %Neg %Avg. Buzz
2026 NAFTA at Stanford01/17/20264100%50%0%71.00
2026 NAFTA at UBC01/17/20262100%50%50%104.00
2025 NAFTA Online02/14/20264100%0%0%89.50
2026 NAFTA at Vanderbilt02/14/20263100%67%0%88.33
2025 NAFTA at Richmond12/20/20252100%50%50%104.00