Datasets › PACS

PACS (Photo-Art-Cartoon-Sketch)

Introduced by Da Li et al. in Deeper, Broader and Artier Domain Generalization1 Jan 2017 archive 2025-07-28

PACS is an image dataset for domain generalization. It consists of four domains, namely Photo (1,670 images), Art Painting (2,048 images), Cartoon (2,344 images) and Sketch (3,929 images). Each domain contains seven categories.

Source: Deep Domain-Adversarial Image Generation for Domain Generalisation Image Source: https://www.researchgate.net/figure/Sample-images-from-PACS-dataset-Each-row-represents-a-domain-and-each-column-represents_fig1_334695033

Benchmarks archive 2025-07-28

All 10 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Domain Generalization PACS SIMPLE+ Average Accuracy 99.0 SIMPLE: Specialized Model-Sample Matching for Domain... microsoft/SeqML 133 Compare
Single-Source Domain Generalization PACS Crafting-Shifts(ResNet18) Accuracy 70.37 Crafting Distribution Shifts for Validation and Training... nikosefth/crafting-shifts 10 Compare
Photo to Rest Generalization PACS Crafting-Shifts(ResNet18) Accuracy 65.85 Crafting Distribution Shifts for Validation and Training... nikosefth/crafting-shifts 8 Compare
Image to sketch recognition PACS Crafting-Shifts(ResNet18) Accuracy 74.13 Crafting Distribution Shifts for Validation and Training... nikosefth/crafting-shifts 7 Compare
Annotated Code Search PACS-CoNaLa Ensemble:USE-tuned+NCS MRR 0.351 Neural Code Search Revisited: Enhancing Code Snippet... nokia/codesearch 4 Compare
Annotated Code Search PACS-SO-DS Ensemble:USE-tuned+NCS MRR 0.323 Neural Code Search Revisited: Enhancing Code Snippet... nokia/codesearch 4 Compare
Annotated Code Search PACS-StaQC-py Ensemble:USE-tuned+NCS MRR 0.126 Neural Code Search Revisited: Enhancing Code Snippet... nokia/codesearch 4 Compare
Source-Free Domain Adaptation PACS SPM Average Accuracy 86.7 Shuffle PatchMix Augmentation with Confidence-Margin... PrasannaPulakurthi/SPM 3 Compare
Unsupervised Domain Adaptation PACS CoVi Average Accuracy 93.52 Contrastive Vicinal Space for Unsupervised Domain Adaptation najaemin92/covi 3 Compare
Domain Adaptation PACS SSGEN Accuracy 88.09 Learning to Generalize across Domains on Single Test Samples zzzx1224/singlesamplegeneralization-iclr2022 1 Compare

Papers archive 2025-07-28

30 shown of 102 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 668. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
Shuffle PatchMix Augmentation with Confidence-Margin Weighted Pseudo-Labels for Enhanced Source-Free Domain Adaptation 1 1 30 May 2025 not harvested
Effective Dual-Region Augmentation for Reduced Reliance on Large Amounts of Labeled Data 1 1 17 Apr 2025 not harvested
QT-DoG: Quantization-aware Training for Domain Generalization 1 2 8 Oct 2024 ran 1 of 1 samples (0 unverified; 1 pointer-only for licence)
Crafting Distribution Shifts for Validation and Training in Single Source Domain Generalization 1 5 29 Sep 2024 not harvested
Soft Prompt Generation for Domain Generalization 1 2 30 Apr 2024 ran 6 of 9 samples (3 unverified)
Rethinking Multi-domain Generalization with A General Learning Objective 2 4 29 Feb 2024 not harvested
Domain Generalization Using Large Pretrained Models with Mixture-of-Adapters 1 1 17 Oct 2023 not harvested
Towards Unified and Effective Domain Generalization 1 1 16 Oct 2023 not harvested
Leveraging Vision-Language Models for Improving Domain Generalization in Image Classification 1 1 12 Oct 2023 ran 12 of 15 samples (3 unverified)
A Sentence Speaks a Thousand Images: Domain Generalization through Distilling CLIP with Language Guidance 1 1 21 Sep 2023 ran 14 of 16 samples (2 unverified; 16 pointer-only for licence)
Weight Averaging Improves Knowledge Distillation under Domain Shift 1 2 20 Sep 2023 not harvested
PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalization 1 3 27 Jul 2023 ran 0 of 1 samples (1 unverified)
Adversarial Bayesian Augmentation for Single-Source Domain Generalization 1 2 18 Jul 2023 ran 4 of 5 samples (1 unverified; 5 pointer-only for licence)
POEM: Polarization of Embeddings for Domain-Invariant Representations 1 1 22 May 2023 ran 9 of 12 samples (3 unverified)
SIMPLE: Specialized Model-Sample Matching for Domain Generalization 1 2 1 May 2023 not harvested
Improved Test-Time Adaptation for Domain Generalization 1 2 10 Apr 2023 ran 0 of 1 samples (1 unverified; 1 pointer-only for licence)
Meta-causal Learning for Single Domain Generalization 0 2 7 Apr 2023 not harvested
VNE: An Effective Method for Improving Deep Representation by Manipulating Eigenvalue Distribution 1 1 4 Apr 2023 not harvested
Progressive Random Convolutions for Single Domain Generalization 0 2 2 Apr 2023 not harvested
Domain-aware Triplet loss in Domain Generalization 1 2 1 Mar 2023 ran 7 of 12 samples (5 unverified)
Center-aware Adversarial Augmentation for Single Domain Generalization 0 2 1 Jan 2023 not harvested
Model Ratatouille: Recycling Diverse Models for Out-of-Distribution Generalization 1 1 20 Dec 2022 not harvested
Domain Generalization by Learning and Removing Domain-specific Features 1 1 14 Dec 2022 ran 2 of 7 samples (5 unverified)
Context-Aware Robust Fine-Tuning 0 1 29 Nov 2022 not harvested
Cross-Domain Ensemble Distillation for Domain Generalization 1 3 25 Nov 2022 not harvested
Meta-DMoE: Adapting to Domain Shift by Meta-Distillation from Mixture-of-Experts 1 1 8 Oct 2022 ran 0 of 7 samples (7 unverified)
Sparse Mixture-of-Experts are Domain Generalizable Learners 2 1 8 Jun 2022 ran 7 of 14 samples (7 unverified; 14 pointer-only for licence)
Dynamic Domain Generalization 1 1 27 May 2022 ran 1 of 1 samples (0 unverified; 1 pointer-only for licence)
CADG: A Model Based on Cross Attention for Domain Generalization 0 1 31 Mar 2022 not harvested
Domain Generalization by Mutual-Information Regularization with Pre-trained Models 1 2 21 Mar 2022 ran 11 of 13 samples (2 unverified)

The full list of 102 is in the JSON twin.

Dataset loaders archive 2025-07-28

1 loader as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

Unknown

Modalities archive 2025-07-28

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • PACS
  • PACS-StaQC-py
  • PACS-SO-DS
  • PACS-CoNaLa

4 variant names, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections