AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Do Image Classifiers Generalize Across Time?

Vaishaal Shankar, Achal Dave, Rebecca Roelofs, Deva Ramanan, Benjamin Recht, Ludwig Schmidt

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 65 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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
add noise/corruption augmentation to ILSVRC training (ablation)ResNet-50 · ImageNet-Vid-Robustaccuracy drop: 15.0 -> 15.6No clear effect
add noise/corruption augmentation to ILSVRC training (ablation)ResNet-50 · YTBB-Robustaccuracy drop: 13.2 -> 16.6Hurt
use an l_inf adversarially robust model (ablation)ResNeXt-101 · ImageNet-Vid-Robustaccuracy drop: 15.0 -> 12.4Mixed
use an l_inf adversarially robust model (ablation)ResNeXt-101 · YTBB-Robustaccuracy drop: 13.2 -> 10.4Mixed
fine-tune on ImageNet-Vid frames (ablation)ResNet-50 · ImageNet-Vid-Robustaccuracy drop: 15.0 -> 15.1No clear effect
fine-tune on Youtube-BB frames (ablation)ResNet-50 · YTBB-Robustaccuracy drop: 13.2 -> 9.4Improved
fine-tune on ImageNet-Vid frames (ablation)ResNet-50 · ImageNet-Vid-Robustperturbed accuracy: 52.5 -> 65.7Improved
train a Faster R-CNN detector instead of a classifier (ablation)ResNet-50 fine-tuned on ImageNet-Vid · ImageNet-Vid-Robustoriginal accuracy: 80.8 -> 77.6Hurt
add bounding-box supervision (detection fine-tuning) (ablation)ResNet-50 · ImageNet-Vid-Robustaccuracy drop: 15.1 -> 12.3No clear effect
switch from detection to localization only (ablation)Faster R-CNN R50 · ImageNet-Vid-RobustmAP drop: 14.0 -> 12.4No clear effect
add human review of frame similarity and labels (ablation)ResNet-152 FT · ImageNet-Vid-Robustaccuracy drop: 16.2 -> 14.4Improved
add human review of frame similarity and labels (ablation)ResNet-152 FT · YTBB-Robustaccuracy drop: 10.0 -> 8.9Improved

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