A Rigorous Behavior Assessment of CNNs Using a Data-Domain Sampling Regime

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Jiang, Shuning, Chao, Wei-Lun, Haehn, Daniel, Pfister, Hanspeter, Chen, Jian
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915507671662592
author Jiang, Shuning
Chao, Wei-Lun
Haehn, Daniel
Pfister, Hanspeter
Chen, Jian
author_facet Jiang, Shuning
Chao, Wei-Lun
Haehn, Daniel
Pfister, Hanspeter
Chen, Jian
contents We present a data-domain sampling regime for quantifying CNNs' graphic perception behaviors. This regime lets us evaluate CNNs' ratio estimation ability in bar charts from three perspectives: sensitivity to training-test distribution discrepancies, stability to limited samples, and relative expertise to human observers. After analyzing 16 million trials from 800 CNNs models and 6,825 trials from 113 human participants, we arrived at a simple and actionable conclusion: CNNs can outperform humans and their biases simply depend on the training-test distance. We show evidence of this simple, elegant behavior of the machines when they interpret visualization images. osf.io/gfqc3 provides registration, the code for our sampling regime, and experimental results.
format Preprint
id arxiv_https___arxiv_org_abs_2507_03866
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Rigorous Behavior Assessment of CNNs Using a Data-Domain Sampling Regime
Jiang, Shuning
Chao, Wei-Lun
Haehn, Daniel
Pfister, Hanspeter
Chen, Jian
Machine Learning
Computer Vision and Pattern Recognition
Human-Computer Interaction
We present a data-domain sampling regime for quantifying CNNs' graphic perception behaviors. This regime lets us evaluate CNNs' ratio estimation ability in bar charts from three perspectives: sensitivity to training-test distribution discrepancies, stability to limited samples, and relative expertise to human observers. After analyzing 16 million trials from 800 CNNs models and 6,825 trials from 113 human participants, we arrived at a simple and actionable conclusion: CNNs can outperform humans and their biases simply depend on the training-test distance. We show evidence of this simple, elegant behavior of the machines when they interpret visualization images. osf.io/gfqc3 provides registration, the code for our sampling regime, and experimental results.
title A Rigorous Behavior Assessment of CNNs Using a Data-Domain Sampling Regime
topic Machine Learning
Computer Vision and Pattern Recognition
Human-Computer Interaction
url https://arxiv.org/abs/2507.03866