평범한 필기장

https://arxiv.org/abs/2403.17377 Self-Rectifying Diffusion Sampling with Perturbed-Attention GuidanceRecent studies have demonstrated that diffusion models are capable of generating high-quality samples, but their quality heavily depends on sampling guidance techniques, such as classifier guidance (CG) and classifier-free guidance (CFG). These techniquesarxiv.org1. Introduction Diffusion Model들은..

https://arxiv.org/abs/2403.18818 ObjectDrop: Bootstrapping Counterfactuals for Photorealistic Object Removal and InsertionDiffusion models have revolutionized image editing but often generate images that violate physical laws, particularly the effects of objects on the scene, e.g., occlusions, shadows, and reflections. By analyzing the limitations of self-supervised approachearxiv.org1. Introduc..

https://arxiv.org/abs/2403.17804 Improving Text-to-Image Consistency via Automatic Prompt Optimization Impressive advances in text-to-image (T2I) generative models have yielded a plethora of high performing models which are able to generate aesthetically appealing, photorealistic images. Despite the progress, these models still struggle to produce images th arxiv.org 1. Introduction 기존의 T2I 모델들은..

주재걸 교수님의 DAVIAN Lab에서 진행하는 computer vision study를 청강하게 되었다. 최신 논문들을 다루는 것 같아서 따라가기 힘들겠지만 최대한 스터디 전에 간단하게 어떤 논문인지 맛보고 스터디 청강을 해야겠다는 생각이 들었다. 그래서 이 스터디에서 읽을 논문들은 최대한 어떤 논문인지 간단하게만 정리해보려고 한다. 이번 주의 논문은 "VAR"이다. https://arxiv.org/abs/2404.02905