AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Self-Supervised Learning Across Domains

Silvia Bucci, Antonio D’Innocente, Yujun Liao, Fabio Maria Carlucci, Barbara Caputo B, Tatiana Tommasi

TPAMI 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Invariance-based self-supervised learning · 80 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 jigsaw self-supervised task to supervised training (ablation)AlexNet DeepAll · PACS DGaccuracy: 71.81 -> 73.05Improved
add rotation self-supervised task to supervised training (ablation)ResNet-18 DeepAll · PACS DGaccuracy: 78.40 -> 81.87Improved
combine jigsaw and rotation instead of jigsaw alone (ablation)AlexNet multi-task · PACS DGaccuracy: 73.05 -> 74.08Improved
combine jigsaw and rotation instead of rotation alone (ablation)ResNet-18 multi-task · PACS DGaccuracy: 81.87 -> 81.41No clear effect
add rotation task to supervised training (ablation)AlexNet DeepAll · VLCS DGaccuracy: 72.49 -> 73.88Improved
combine jigsaw and rotation instead of rotation alone (ablation)AlexNet multi-task · VLCS DGaccuracy: 73.88 -> 73.15Hurt
use rotation instead of jigsaw as auxiliary task (ablation)ResNet-18 multi-task · Office-Home DGaccuracy: 61.20 -> 63.53Improved
use whole-image jigsaw pretext (J-AlexNet) instead of patch-based J-CFN (ablation)self-supervised pretraining · PACS DGaccuracy: 59.41 -> 51.48Hurt
add jigsaw auxiliary task to the multi-branch CFN architecture (ablation)C-CFN DeepAll · PACS DGaccuracy: 62.66 -> 64.89Improved
add self-supervised task on target data (ablation)Jigsaw+Rotation ResNet-18 · PACS multi-source DAaccuracy: 83.73 -> 88.66Improved
add entropy loss (ablation)Jigsaw+Rotation ResNet-18 · PACS multi-source DAaccuracy: 85.23 -> 88.66Improved
add self-supervised task on source data (ablation)Jigsaw ResNet-50 · Office-Home single-source DAaccuracy: 62.06 -> 62.17No 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