VLCS

The VLCS dataset is a benchmark that contains images from four different domains (Pascal VOC, LabelMe, Caltech 101, and SUN) and is used to evaluate the performance of domain generalization methods.

Papers using VLCS (4)

Federated Domain Generalization with Data-free On-server Matching Gradient2025 · 1 citesFedSDAF: Leveraging Source Domain Awareness for Enhanced Federated Domain Generalization2025Federated Domain Generalization with Label Smoothing and Balanced Decentralized Training2024Hypernetwork-Driven Model Fusion for Federated Domain Generalization2024