Best of WACV – May 29

Given a small number of images of a subject, personalized image generation techniques can fine-tune large pre-trained text-to-image diffusion models to generate images of the subject in novel contexts, conditioned on text prompts. In doing so, a trade-off is made between prompt fidelity, subject fidelity and diversity. As the pre-trained model is fine-tuned, earlier checkpoints synthesize images with low subject fidelity but high prompt fidelity and diversity. In contrast, later checkpoints generate images with low prompt fidelity and diversity but high subject fidelity. This inherent trade-off limits the prompt fidelity, subject fidelity and diversity of generated images. In this work, we propose DreamBlend to combine the prompt fidelity from earlier checkpoints and the subject fidelity from later checkpoints during inference. We perform a cross attention guided image synthesis from a later checkpoint, guided by an image generated by an earlier checkpoint, for the same prompt. This enables generation of images with better subject fidelity, prompt fidelity and diversity on challenging prompts, outperforming state-of-the-art fine-tuning methods.

Paper: DreamBlend: Advancing Personalized Fine-tuning of Text-to-Image Diffusion Models

Robust multi-class anomaly detection under domain shift is a fundamental challenge in real-world scenarios, where detectors should distinguish different types of anomalies despite significant distribution shifts. Traditional approaches often struggle to generalize across domains and handle inter-class interference. ROADS addresses these limitations through a prompt-driven framework that combines a hierarchical class-aware prompt mechanism with a domain adapter to jointly encode discriminative, class-specific prompts and learn domain-invariant representations. Extensive evaluations on the MVTec-AD and VISA datasets show that ROADS achieves superior performance in both anomaly detection and localization, particularly in out-of-distribution settings.

Paper: ROADS: Robust Prompt-driven Multi-Class Anomaly Detection under Domain Shift