BiRQA: Bidirectional Robust Quality Assessment for Images

ICLR 2026 Conference Submission19092 Authors

Keywords: image quality assessment, adversarial training, adversarial attack, robustness

TL;DR: BiRQA is a full-reference IQA model. It’s fast (≈15 FPS on 1080p), accurate (state-of-the-art on 5 FR-IQA benchmarks), and robust thanks to the proposed Anchored Adversarial Training.

Abstract: Full-Reference image quality assessment (FR IQA) is important for image compression, restoration and generative modeling, yet current neural metrics remain slow and vulnerable to adversarial perturbations. We present BiRQA, a compact FR IQA metric model that processes four fast complementary features within a bidirectional multiscale pyramid. A bottom-up attention module injects fine-scale cues into coarse levels through an uncertainty-aware gate, while a top-down cross-gating block routes semantic context back to high resolution. To enhance robustness, we introduce Anchored Adversarial Training, a theoretically grounded strategy that uses clean "anchor" samples and a ranking loss to bound pointwise prediction error under attacks. On five public FR IQA benchmarks BiRQA outperforms or matches the previous state of the art (SOTA) while running $\sim3\times$ faster than previous SOTA models. Under unseen white-box attacks it lifts SROCC from 0.30-0.57 to 0.60-0.84 on KADID-10k, demonstrating substantial robustness gains. To our knowledge, BiRQA is the only FR IQA model combining competitive accuracy with real-time throughput and strong adversarial resilience.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 19092