APEX: Assumption-free Projection-based Embedding eXamination Metric for Image Quality Assessment

TL;DR

APEX introduces a novel, assumption-free image quality assessment metric using Sliced Wasserstein Distance with open-vocabulary models, demonstrating superior robustness and stability over traditional methods.

Contribution

It proposes a new evaluation framework that overcomes parametric limitations and feature bottlenecks of existing metrics, leveraging assumption-free similarity measures and open-vocabulary models.

Findings

APEX outperforms traditional metrics in robustness to visual degradations.

APEX exhibits high stability across different datasets and domains.

Theoretical and empirical evidence supports APEX's scalability to high-dimensional spaces.

Abstract

As generative models achieve unprecedented visual quality, the gold standard for image evaluation remains traditional feature-distribution metrics (e.g., FID). However, these metrics are provably hindered by the closed-vocabulary bottleneck of outdated features and the assumptive bias of rigid parametric formulations. Recent alternatives exploit modern backbones to solve the feature bottleneck, yet continue to suffer from parametric limitations. To close this gap, we introduce APEX (Assumption-free Projection-based Embedding eXamination), a novel evaluation framework leveraging the Sliced Wasserstein Distance as a mathematically grounded, assumption-free similarity measure. APEX inherits effective scalability to high-dimensional spaces, as we prove with theoretical and empirical evidences. Moreover, APEX is embedding-agnostic and uses two open-vocabulary foundation models, CLIP and DINOv2, as feature extractors. Benchmarking APEX against established baselines reveals superior robustness to visual degradations. Additionally, we show that APEX metrics exhibit intra- and cross-dataset stability, ensuring highly stable evaluations on out-of-domain datasets.