Generative models often use human evaluations to measure the perceived quality of their outputs. Automated metrics are noisy indirect proxies, because they rely on heuristics or pretrained embeddings. However, up until now, direct human evaluation strategies have been ad-hoc, neither standardized nor validated. Our work establishes a gold standard human benchmark for generative realism. We construct Human eYe Perceptual Evaluation (HYPE) a human benchmark that is (1) grounded in psychophysics research in perception, (2) reliable across different sets of randomly sampled outputs from a model, (3) able to produce separable model performances, and (4) efficient in cost and time. We introduce two variants: one that measures visual perception under adaptive time constraints to determine the threshold at which a model's outputs appear real (e.g. $250$ms), and the other a less expensive variant that measures human error rate on fake and real images sans time constraints. We test HYPE across six state-of-the-art generative adversarial networks and two sampling techniques on conditional and unconditional image generation using four datasets: CelebA, FFHQ, CIFAR-10, and ImageNet. We find that HYPE can track model improvements across training epochs, and we confirm via bootstrap sampling that HYPE rankings are consistent and replicable.
Author Information
Sharon Zhou (Stanford University)
Mitchell Gordon (Stanford University)
Ranjay Krishna (Stanford University)
Austin Narcomey (Stanford University)
Li Fei-Fei (Stanford University)
Michael Bernstein (Stanford University)
Related Events (a corresponding poster, oral, or spotlight)
-
2019 Poster: HYPE: A Benchmark for Human eYe Perceptual Evaluation of Generative Models »
Tue Dec 10th 05:30 -- 07:30 PM Room East Exhibition Hall B + C
More from the Same Authors
-
2019 Poster: Regression Planning Networks »
Danfei Xu · Roberto Martín-Martín · De-An Huang · Yuke Zhu · Silvio Savarese · Li Fei-Fei -
2018 Poster: Learning to Play With Intrinsically-Motivated, Self-Aware Agents »
Nick Haber · Damian Mrowca · Stephanie Wang · Li Fei-Fei · Daniel Yamins -
2018 Poster: Learning to Decompose and Disentangle Representations for Video Prediction »
Jun-Ting Hsieh · Bingbin Liu · De-An Huang · Li Fei-Fei · Juan Carlos Niebles -
2018 Poster: Flexible neural representation for physics prediction »
Damian Mrowca · Chengxu Zhuang · Elias Wang · Nick Haber · Li Fei-Fei · Josh Tenenbaum · Daniel Yamins -
2017 Poster: Label Efficient Learning of Transferable Representations acrosss Domains and Tasks »
Zelun Luo · Yuliang Zou · Judy Hoffman · Li Fei-Fei -
2014 Poster: Deep Fragment Embeddings for Bidirectional Image Sentence Mapping »
Andrej Karpathy · Armand Joulin · Li Fei-Fei -
2013 Workshop: Crowdsourcing: Theory, Algorithms and Applications »
Jennifer Wortman Vaughan · Greg Stoddard · Chien-Ju Ho · Adish Singla · Michael Bernstein · Devavrat Shah · Arpita Ghosh · Evgeniy Gabrilovich · Denny Zhou · Nikhil Devanur · Xi Chen · Alexander Ihler · Qiang Liu · Genevieve Patterson · Ashwinkumar Badanidiyuru Varadaraja · Hossein Azari Soufiani · Jacob Whitehill -
2012 Workshop: Big Data Meets Computer Vision: First International Workshop on Large Scale Visual Recognition and Retrieval »
Jia Deng · Samy Bengio · Yuanqing Lin · Li Fei-Fei -
2012 Poster: Shifting Weights: Adapting Object Detectors from Image to Video »
Kevin Tang · Vignesh Ramanathan · Li Fei-Fei · Daphne Koller -
2012 Demonstration: EVA: Engine for Visual Annotation »
Jia Deng · Joanathan Krause · Zhiheng Huang · Alexander C Berg · Li Fei-Fei -
2011 Poster: Fast and Balanced: Efficient Label Tree Learning for Large Scale Object Recognition »
Jia Deng · Sanjeev Satheesh · Alexander C Berg · Li Fei-Fei -
2011 Poster: Large-Scale Category Structure Aware Image Categorization »
Bin Zhao · Li Fei-Fei · Eric Xing -
2010 Session: Oral Session 10 »
Li Fei-Fei -
2010 Poster: Large Margin Learning of Upstream Scene Understanding Models »
Jun Zhu · Li-Jia Li · Li Fei-Fei · Eric Xing -
2010 Poster: Object Bank: A High-Level Image Representation for Scene Classification & Semantic Feature Sparsification »
Li-Jia Li · Hao Su · Eric Xing · Li Fei-Fei
