Beyond First-Order Tweedie: Solving Inverse Problems using Latent Diffusion
원문 초록
Sampling from the posterior distribution poses a major computational challenge in solving inverse problems using latent diffusion models. Common methods rely on Tweedie's first-order moments, which are known to induce a quality-limiting bias. Existing second-order approximations are impractical due to prohibitive computational costs, making standard reverse diffusion processes intractable for posterior sampling. This paper introduces Second-order Tweedie sampler from Surrogate Loss (STSL), a novel sampler that offers efficiency comparable to first-order Tweedie with a tractable reverse process using second-order approximation. Our theoretical results reveal that the second-order approximation is lower bounded by our surrogate loss that only requires $O(1)$ compute using the trace of the Hessian, and by the lower bound we derive a new drift term to make the reverse process tractable. Our method surpasses SoTA solvers PSLD and P2L, achieving 4X and 8X reduction in neural function evaluations, respectively, while notably enhancing sampling quality on FFHQ, ImageNet, and COCO benchmarks. In addition, we show STSL extends to text-guided image editing and addresses residual distortions present from corrupted images in leading text-guided image editing methods. To our best knowledge, this is the first work to offer an efficient second-order approximation in solving inverse problems using latent diffusion and editing real-world images with corruptions.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
PairEdit: Learning Semantic Variations for Exemplar-based Image Editing
Highly Compressed Tokenizer Can Generate Without Training
GIE-Bench: Towards Grounded Evaluation for Text-Guided Image Editing
Towards Scalable Human-aligned Benchmark for Text-guided Image Editing
Training-Free Text-Guided Image Editing with Visual Autoregressive Model
LOCATEdit: Graph Laplacian Optimized Cross Attention for Localized Text-Guided Image Editing
토론
이 논문과 연결된 포럼 글이 아직 없습니다.