Robustness of Vision Foundation Models to Common Perturbations

ArXi:2604.14973v1 Announce Type: cross A vision foundation model outputs an embedding vector for an image, which can be affected by common editing operations (e.g., JPEG compression, brightness, contrast adjustments). These common perturbations alter embedding vectors and may impact the performance of downstream tasks using these embeddings. In this work, we present the first systematic study on foundation models' robustness to such perturbations.