CVMar 13

DiveUp: Learning Feature Upsampling from Diverse Vision Foundation Models

arXiv:2603.1357178.9h-index: 4Has Code
AI Analysis

This addresses a fundamental limitation in enhancing vision foundation models for pixel-level understanding tasks, offering a unified solution that improves accuracy without per-model retraining.

The paper tackles the problem of feature upsampling in vision foundation models by proposing DiveUp, a framework that uses multi-model relational guidance instead of relying on single-model features, achieving state-of-the-art performance across various dense prediction tasks.

Recently, feature upsampling has gained increasing attention owing to its effectiveness in enhancing vision foundation models (VFMs) for pixel-level understanding tasks. Existing methods typically rely on high-resolution features from the same foundation model to achieve upsampling via self-reconstruction. However, relying solely on intra-model features forces the upsampler to overfit to the source model's inherent location misalignment and high-norm artifacts. To address this fundamental limitation, we propose DiveUp, a novel framework that breaks away from single-model dependency by introducing multi-VFM relational guidance. Instead of naive feature fusion, DiveUp leverages diverse VFMs as a panel of experts, utilizing their structural consensus to regularize the upsampler's learning process, effectively preventing the propagation of inaccurate spatial structures from the source model. To reconcile the unaligned feature spaces across different VFMs, we propose a universal relational feature representation, formulated as a local center-of-mass (COM) field, that extracts intrinsic geometric structures, enabling seamless cross-model interaction. Furthermore, we introduce a spikiness-aware selection strategy that evaluates the spatial reliability of each VFM, effectively filtering out high-norm artifacts to aggregate guidance from only the most reliable expert at each local region. DiveUp is a unified, encoder-agnostic framework; a jointly-trained model can universally upsample features from diverse VFMs without requiring per-model retraining. Extensive experiments demonstrate that DiveUp achieves state-of-the-art performance across various downstream dense prediction tasks, validating the efficacy of multi-expert relational guidance. Our code and models are available at: https://github.com/Xiaoqiong-Liu/DiveUp

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes