평범한 필기장

paper : https://arxiv.org/abs/2410.12557 One Step Diffusion via Shortcut ModelsDiffusion models and flow-matching models have enabled generating diverse and realistic images by learning to transfer noise to data. However, sampling from these models involves iterative denoising over many neural network passes, making generation slow aarxiv.orgAbstract본 논문은 shortcut model을 제안한다. 이는 single network를..

Paper : https://arxiv.org/abs/2411.00113 A Geometric Framework for Understanding Memorization in Generative ModelsAs deep generative models have progressed, recent work has shown them to be capable of memorizing and reproducing training datapoints when deployed. These findings call into question the usability of generative models, especially in light of the legal andarxiv.orgAbstract본 논문은 memori..

Paper : https://arxiv.org/abs/2406.03537 A Geometric View of Data Complexity: Efficient Local Intrinsic Dimension Estimation with Diffusion ModelsHigh-dimensional data commonly lies on low-dimensional submanifolds, and estimating the local intrinsic dimension (LID) of a datum -- i.e. the dimension of the submanifold it belongs to -- is a longstanding problem. LID can be understood as the number ..

Paper : https://arxiv.org/abs/2404.04650 InitNO: Boosting Text-to-Image Diffusion Models via Initial Noise OptimizationRecent strides in the development of diffusion models, exemplified by advancements such as Stable Diffusion, have underscored their remarkable prowess in generating visually compelling images. However, the imperative of achieving a seamless alignment betwearxiv.orgAbstract 모든 ra..

Paper : https://arxiv.org/abs/2411.16738 Classifier-Free Guidance inside the Attraction Basin May Cause MemorizationDiffusion models are prone to exactly reproduce images from the training data. This exact reproduction of the training data is concerning as it can lead to copyright infringement and/or leakage of privacy-sensitive information. In this paper, we present aarxiv.orgAbstract해결하려는 문제 D..

Paper : https://arxiv.org/abs/2407.21720 Detecting, Explaining, and Mitigating Memorization in Diffusion ModelsRecent breakthroughs in diffusion models have exhibited exceptional image-generation capabilities. However, studies show that some outputs are merely replications of training data. Such replications present potential legal challenges for model owners, espearxiv.orgAbstract문제 : 생성모델의 몇 o..

Paper : https://arxiv.org/abs/2309.06380 InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image GenerationDiffusion models have revolutionized text-to-image generation with its exceptional quality and creativity. However, its multi-step sampling process is known to be slow, often requiring tens of inference steps to obtain satisfactory results. Previous attemparxiv.orgAbstr..

Paper : https://arxiv.org/abs/2304.08465 MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and EditingDespite the success in large-scale text-to-image generation and text-conditioned image editing, existing methods still struggle to produce consistent generation and editing results. For example, generation approaches usually fail to synthesize multiple imaarxiv.o..

Paper : https://arxiv.org/abs/2310.01506 Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of CodeText-guided diffusion models have revolutionized image generation and editing, offering exceptional realism and diversity. Specifically, in the context of diffusion-based editing, where a source image is edited according to a target prompt, the process comarxiv.orgProject Page : https:..

Paper : https://arxiv.org/abs/2403.12002 DreamMotion: Space-Time Self-Similar Score Distillation for Zero-Shot Video EditingText-driven diffusion-based video editing presents a unique challenge not encountered in image editing literature: establishing real-world motion. Unlike existing video editing approaches, here we focus on score distillation sampling to circumvent the stanarxiv.orgProject P..