BiRQA: Bidirectional Robust Quality Assessment for Images
This addresses the need for efficient and robust image quality assessment in applications like compression and generative modeling, offering a novel combination of accuracy, speed, and resilience, though it is incremental in improving existing FR IQA methods.
The paper tackles the problem of slow and vulnerable full-reference image quality assessment (FR IQA) metrics by introducing BiRQA, a compact model that processes fast complementary features with a bidirectional multiscale pyramid and anchored adversarial training, achieving state-of-the-art performance while running ~3x faster and improving robustness under attacks, with SROCC gains from 0.30-0.57 to 0.60-0.84 on KADID-10k.
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 ~3x 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.