2025/03/11 by Chen, Hesen, Junyan Wang, Wang, Junyan +4 · 4 citations
Computer Science · Physics and Astronomy · #Adversarial Robustness in Machine Learning #Adversarial system #Coherence (philosophical gambling strategy) #Computer Vision and Pattern Recognition (cs.CV) #Consistency (knowledge bases) #Convergence (economics) #Encoder #FOS: Computer and information sciences #Model Reduction and Neural Networks #Pattern recognition (psychology) #Representation (politics)
paper · pdf · doi:10.48550/arxiv.2503.08253
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/03/11 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Modern diffusion models encounter a fundamental trade-off between training efficiency and generation quality. While existing representation alignment methods, such as REPA, accelerate convergence through patch-wise alignment, they often fail to capture structural relationships within visual representations and ensure global distribution consistency between pretrained encoders and denoising networks. To address these limitations, we introduce SARA, a hierarchical alignment framework that enforces multi-level representation constraints: (1) patch-wise alignment to preserve local semantic details, (2) autocorrelation matrix alignment to maintain structural consistency within representations, and (3) adversarial distribution alignment to mitigate global representation discrepancies. Unlike previous approaches, SARA explicitly models both intra-representation correlations via self-similarity matrices and inter-distribution coherence via adversarial alignment, enabling comprehensive alignment across local and global scales. Experiments on ImageNet-256 show that SARA achieves an FID of 1.36 while converging twice as fast as REPA, surpassing recent state-of-the-art image generation methods. This work establishes a systematic paradigm for optimizing diffusion training through hierarchical representation alignment.