AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Adaptive Testing of Computer Vision Models

Irena Gao, Gabriel Ilharco, Scott Lundberg, Marco Tulio Ribeiro

ICCV 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Human-like visual cognition in VLMs · 35 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use adaptive test suggestions (hill-climb on previous failures) instead of topic-name-only retrieval, user study (ablation)AdaVision · User study (classification)failures found: d=0.588, p < 0.05Improved
use adaptive test suggestions instead of non-adaptive retrieval (ablation)AdaVision · User study (object detection)failures found: d=0.882, p < 0.005Improved
use adaptive test suggestions instead of non-adaptive retrieval (ablation)AdaVision · User study (image captioning)failures found: d=0.967, p < 0.05Improved
use AdaVision instead of NonAdaptive for perceived difficulty (ablation)AdaVision · User studyperceived difficulty (1-5, lower easier): 4.10 ± 0.91 -> 3.05 ± 1.07Improved
use human-in-the-loop AdaVision topics instead of Domino automatic slicesViT-H/14 · ImageNet categories, new retrievalsaverage failure rate: 8.6 -> 28.47Improved
transfer AdaVision topics found on ViT-H/14 to ResNet-50ResNet-50 · ImageNet categories, new retrievalsaverage failure rate: 25.45 -> 56.93Improved
finetune on AdaVision-found tests instead of generic retrieved images (ablation)ViT-H/14 · AdaVision treatment topics (LAION-5B held-out)accuracy: 82.5 -> 91.2Improved
finetune on AdaVision testsViT-H/14 · ImageNet valaccuracy: 88.4 -> 88.4No clear effect
finetune on AdaVision testsViT-H/14 · Five ImageNet OOD sets, treatment classesaccuracy: 78.0 -> 84.0Improved
finetune on AdaVision testsViT-H/14 · Control topicsaccuracy: 91.3 -> 91.9No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy