MuRF: Unlocking the Multi-Scale Potential of Vision Foundation Models
This addresses a fundamental problem in computer vision by enhancing visual representations for various tasks, though it is incremental as it builds on existing models without architectural changes.
The paper tackles the limitation of vision foundation models being restricted to single-scale inference by proposing MuRF, a training-free method that processes images at multiple resolutions and fuses features, achieving improvements such as a 2.1% gain in mIoU on ADE20K segmentation and 1.5% in accuracy on ImageNet classification.
Vision Foundation Models (VFMs) have become the cornerstone of modern computer vision, offering robust representations across a wide array of tasks. While recent advances allow these models to handle varying input sizes during training, inference typically remains restricted to a single, fixed scale. This prevalent single-scale paradigm overlooks a fundamental property of visual perception: varying resolutions offer complementary inductive biases, where low-resolution views excel at global semantic recognition and high-resolution views are essential for fine-grained refinement. In this work, we propose Multi-Resolution Fusion (MuRF), a simple yet universally effective strategy to harness this synergy at inference time. Instead of relying on a single view, MuRF constructs a unified representation by processing an image at multiple resolutions through a frozen VFM and fusing the resulting features. The universality of MuRF is its most compelling attribute. It is not tied to a specific architecture, serving instead as a fundamental, training-free enhancement to visual representation. We empirically validate this by applying MuRF to a broad spectrum of critical computer vision tasks across multiple distinct VFM families - primarily DINOv2, but also demonstrating successful generalization to contrastive models like SigLIP2.