PyTorchKR Papers

최신 논문

총 7,085편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2406.01956

Enhance Image-to-Image Generation with LLaVA-generated Prompts

This paper presents a novel approach to enhance image-to-image generation by leveraging the multimodal capabilities of the Large Language and Vision Assistant (LLaVA). We propose a framework where LLaVA analyzes input images and generates textual descriptions, hereinafter LLaVA-generated prompts. These prompts, along with the original image, are fed into the image-to-image generation pipeline. This enriched representation guides the generation process towards outputs that exhibit a stronger resemblance to the input image. Extensive experiments demonstrate the effectiveness of LLaVA-generated prompts in promoting image similarity. We observe a significant improvement in the visual coherence between the generated and input images compared to traditional methods. Future work will explore fine-tuning LLaVA prompts for increased control over the creative process. By providing more specific details within the prompts, we aim to achieve a delicate balance between faithfulness to the original image and artistic expression in the generated outputs.

arXiv:2406.01954

Plug-and-Play Diffusion Distillation

Diffusion models have shown tremendous results in image generation. However, due to the iterative nature of the diffusion process and its reliance on classifier-free guidance, inference times are slow. In this paper, we propose a new distillation approach for guided diffusion models in which an external lightweight guide model is trained while the original text-to-image model remains frozen. We show that our method reduces the inference computation of classifier-free guided latent-space diffusion models by almost half, and only requires 1\% trainable parameters of the base model. Furthermore, once trained, our guide model can be applied to various fine-tuned, domain-specific versions of the base diffusion model without the need for additional training: this "plug-and-play" functionality drastically improves inference computation while maintaining the visual fidelity of generated images. Empirically, we show that our approach is able to produce visually appealing results and achieve a comparable FID score to the teacher with as few as 8 to 16 steps.

arXiv:2406.00777공식 코드 · Papers with Code

Diffusion Features to Bridge Domain Gap for Semantic Segmentation

Pre-trained diffusion models have demonstrated remarkable proficiency in synthesizing images across a wide range of scenarios with customizable prompts, indicating their effective capacity to capture universal features. Motivated by this, our study delves into the utilization of the implicit knowledge embedded within diffusion models to address challenges in cross-domain semantic segmentation. This paper investigates the approach that leverages the sampling and fusion techniques to harness the features of diffusion models efficiently. We propose DIffusion Feature Fusion (DIFF) as a backbone use for extracting and integrating effective semantic representations through the diffusion process. By leveraging the strength of text-to-image generation capability, we introduce a new training framework designed to implicitly learn posterior knowledge from it. Through rigorous evaluation in the contexts of domain generalization semantic segmentation, we establish that our methodology surpasses preceding approaches in mitigating discrepancies across distinct domains and attains the state-of-the-art (SOTA) benchmark.

arXiv:2406.01592

Text-guided Controllable Mesh Refinement for Interactive 3D Modeling

We propose a novel technique for adding geometric details to an input coarse 3D mesh guided by a text prompt. Our method is composed of three stages. First, we generate a single-view RGB image conditioned on the input coarse geometry and the input text prompt. This single-view image generation step allows the user to pre-visualize the result and offers stronger conditioning for subsequent multi-view generation. Second, we use our novel multi-view normal generation architecture to jointly generate six different views of the normal images. The joint view generation reduces inconsistencies and leads to sharper details. Third, we optimize our mesh with respect to all views and generate a fine, detailed geometry as output. The resulting method produces an output within seconds and offers explicit user control over the coarse structure, pose, and desired details of the resulting 3D mesh.

arXiv:2406.01561공식 코드 · Papers with Code

Long and Short Guidance in Score identity Distillation for One-Step Text-to-Image Generation

Diffusion-based text-to-image generation models trained on extensive text-image pairs have shown the capacity to generate photorealistic images consistent with textual descriptions. However, a significant limitation of these models is their slow sample generation, which requires iterative refinement through the same network. In this paper, we enhance Score identity Distillation (SiD) by developing long and short classifier-free guidance (LSG) to efficiently distill pretrained Stable Diffusion models without using real training data. SiD aims to optimize a model-based explicit score matching loss, utilizing a score-identity-based approximation alongside the proposed LSG for practical computation. By training exclusively with fake images synthesized with its one-step generator, SiD equipped with LSG rapidly improves FID and CLIP scores, achieving state-of-the-art FID performance while maintaining a competitive CLIP score. Specifically, its data-free distillation of Stable Diffusion 1.5 achieves a record low FID of 8.15 on the COCO-2014 validation set, with a CLIP score of 0.304 at an LSG scale of 1.5, and an FID of 9.56 with a CLIP score of 0.313 at an LSG scale of 2. Our code and distilled one-step text-to-image generators are available at https://github.com/mingyuanzhou/SiD-LSG.

arXiv:2406.01355

Differentially Private Fine-Tuning of Diffusion Models

The integration of Differential Privacy (DP) with diffusion models (DMs) presents a promising yet challenging frontier, particularly due to the substantial memorization capabilities of DMs that pose significant privacy risks. Differential privacy offers a rigorous framework for safeguarding individual data points during model training, with Differential Privacy Stochastic Gradient Descent (DP-SGD) being a prominent implementation. Diffusion method decomposes image generation into iterative steps, theoretically aligning well with DP's incremental noise addition. Despite the natural fit, the unique architecture of DMs necessitates tailored approaches to effectively balance privacy-utility trade-off. Recent developments in this field have highlighted the potential for generating high-quality synthetic data by pre-training on public data (i.e., ImageNet) and fine-tuning on private data, however, there is a pronounced gap in research on optimizing the trade-offs involved in DP settings, particularly concerning parameter efficiency and model scalability. Our work addresses this by proposing a parameter-efficient fine-tuning strategy optimized for private diffusion models, which minimizes the number of trainable parameters to enhance the privacy-utility trade-off. We empirically demonstrate that our method achieves state-of-the-art performance in DP synthesis, significantly surpassing previous benchmarks on widely studied datasets (e.g., with only 0.47M trainable parameters, achieving a more than 35% improvement over the previous state-of-the-art with a small privacy budget on the CelebA-64 dataset). Anonymous codes available at https://anonymous.4open.science/r/DP-LORA-F02F.

arXiv:2406.02347공식 코드 · Papers with Code

Flash Diffusion: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation

In this paper, we propose an efficient, fast, and versatile distillation method to accelerate the generation of pre-trained diffusion models: Flash Diffusion. The method reaches state-of-the-art performances in terms of FID and CLIP-Score for few steps image generation on the COCO2014 and COCO2017 datasets, while requiring only several GPU hours of training and fewer trainable parameters than existing methods. In addition to its efficiency, the versatility of the method is also exposed across several tasks such as text-to-image, inpainting, face-swapping, super-resolution and using different backbones such as UNet-based denoisers (SD1.5, SDXL) or DiT (Pixart-$\alpha$), as well as adapters. In all cases, the method allowed to reduce drastically the number of sampling steps while maintaining very high-quality image generation. The official implementation is available at https://github.com/gojasper/flash-diffusion.

arXiv:2406.01970

The Crystal Ball Hypothesis in diffusion models: Anticipating object positions from initial noise

Diffusion models have achieved remarkable success in text-to-image generation tasks; however, the role of initial noise has been rarely explored. In this study, we identify specific regions within the initial noise image, termed trigger patches, that play a key role for object generation in the resulting images. Notably, these patches are ``universal'' and can be generalized across various positions, seeds, and prompts. To be specific, extracting these patches from one noise and injecting them into another noise leads to object generation in targeted areas. We identify these patches by analyzing the dispersion of object bounding boxes across generated images, leading to the development of a posterior analysis technique. Furthermore, we create a dataset consisting of Gaussian noises labeled with bounding boxes corresponding to the objects appearing in the generated images and train a detector that identifies these patches from the initial noise. To explain the formation of these patches, we reveal that they are outliers in Gaussian noise, and follow distinct distributions through two-sample tests. Finally, we find the misalignment between prompts and the trigger patch patterns can result in unsuccessful image generations. The study proposes a reject-sampling strategy to obtain optimal noise, aiming to improve prompt adherence and positional diversity in image generation.

arXiv:2406.02507공식 코드 · Papers with Code

Guiding a Diffusion Model with a Bad Version of Itself

The primary axes of interest in image-generating diffusion models are image quality, the amount of variation in the results, and how well the results align with a given condition, e.g., a class label or a text prompt. The popular classifier-free guidance approach uses an unconditional model to guide a conditional model, leading to simultaneously better prompt alignment and higher-quality images at the cost of reduced variation. These effects seem inherently entangled, and thus hard to control. We make the surprising observation that it is possible to obtain disentangled control over image quality without compromising the amount of variation by guiding generation using a smaller, less-trained version of the model itself rather than an unconditional model. This leads to significant improvements in ImageNet generation, setting record FIDs of 1.01 for 64x64 and 1.25 for 512x512, using publicly available networks. Furthermore, the method is also applicable to unconditional diffusion models, drastically improving their quality.

arXiv:2406.01278공식 코드 · Papers with Code

fruit-SALAD: A Style Aligned Artwork Dataset to reveal similarity perception in image embeddings

The notion of visual similarity is essential for computer vision, and in applications and studies revolving around vector embeddings of images. However, the scarcity of benchmark datasets poses a significant hurdle in exploring how these models perceive similarity. Here we introduce Style Aligned Artwork Datasets (SALADs), and an example of fruit-SALAD with 10,000 images of fruit depictions. This combined semantic category and style benchmark comprises 100 instances each of 10 easy-to-recognize fruit categories, across 10 easy distinguishable styles. Leveraging a systematic pipeline of generative image synthesis, this visually diverse yet balanced benchmark demonstrates salient differences in semantic category and style similarity weights across various computational models, including machine learning models, feature extraction algorithms, and complexity measures, as well as conceptual models for reference. This meticulously designed dataset offers a controlled and balanced platform for the comparative analysis of similarity perception. The SALAD framework allows the comparison of how these models perform semantic category and style recognition task to go beyond the level of anecdotal knowledge, making it robustly quantifiable and qualitatively interpretable.

arXiv:2406.01125

$Δ$-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers

Diffusion models are widely recognized for generating high-quality and diverse images, but their poor real-time performance has led to numerous acceleration works, primarily focusing on UNet-based structures. With the more successful results achieved by diffusion transformers (DiT), there is still a lack of exploration regarding the impact of DiT structure on generation, as well as the absence of an acceleration framework tailored to the DiT architecture. To tackle these challenges, we conduct an investigation into the correlation between DiT blocks and image generation. Our findings reveal that the front blocks of DiT are associated with the outline of the generated images, while the rear blocks are linked to the details. Based on this insight, we propose an overall training-free inference acceleration framework $\Delta$-DiT: using a designed cache mechanism to accelerate the rear DiT blocks in the early sampling stages and the front DiT blocks in the later stages. Specifically, a DiT-specific cache mechanism called $\Delta$-Cache is proposed, which considers the inputs of the previous sampling image and reduces the bias in the inference. Extensive experiments on PIXART-$\alpha$ and DiT-XL demonstrate that the $\Delta$-DiT can achieve a $1.6\times$ speedup on the 20-step generation and even improves performance in most cases. In the scenario of 4-step consistent model generation and the more challenging $1.12\times$ acceleration, our method significantly outperforms existing methods. Our code will be publicly available.

arXiv:2405.17456공식 코드 · Papers with Code

Generalized Compressed Sensing for Image Reconstruction with Diffusion Probabilistic Models

We examine the problem of selecting a small set of linear measurements for reconstructing high-dimensional signals. Well-established methods for optimizing such measurements include principal component analysis (PCA), independent component analysis (ICA) and compressed sensing (CS) based on random projections, all of which rely on axis- or subspace-aligned statistical characterization of the signal source. However, many naturally occurring signals, including photographic images, contain richer statistical structure. To exploit such structure, we introduce a general method for obtaining an optimized set of linear measurements for efficient image reconstruction, where the signal statistics are expressed by the prior implicit in a neural network trained to perform denoising (known as a ``diffusion model''). We demonstrate that the optimal measurements derived for two natural image datasets differ from those of PCA, ICA, or CS, and result in substantially lower mean squared reconstruction error. Interestingly, the marginal distributions of the measurement values are asymmetrical (skewed), substantially more so than those of previous methods. We also find that optimizing with respect to perceptual loss, as quantified by structural similarity (SSIM), leads to measurements different from those obtained when optimizing for MSE. Our results highlight the importance of incorporating the specific statistical regularities of natural signals when designing effective linear measurements.

arXiv:2405.16119

Method and Software Tool for Generating Artificial Databases of Biomedical Images Based on Deep Neural Networks

A wide variety of biomedical image data, as well as methods for generating training images using basic deep neural networks, were analyzed. Additionally, all platforms for creating images were analyzed, considering their characteristics. The article develops a method for generating artificial biomedical images based on GAN. GAN architecture has been developed for biomedical image synthesis. The data foundation and module for generating training images were designed and implemented in a software system. A comparison of the generated image database with known databases was made.

arXiv:2405.21048

Kaleido Diffusion: Improving Conditional Diffusion Models with Autoregressive Latent Modeling

Diffusion models have emerged as a powerful tool for generating high-quality images from textual descriptions. Despite their successes, these models often exhibit limited diversity in the sampled images, particularly when sampling with a high classifier-free guidance weight. To address this issue, we present Kaleido, a novel approach that enhances the diversity of samples by incorporating autoregressive latent priors. Kaleido integrates an autoregressive language model that encodes the original caption and generates latent variables, serving as abstract and intermediary representations for guiding and facilitating the image generation process. In this paper, we explore a variety of discrete latent representations, including textual descriptions, detection bounding boxes, object blobs, and visual tokens. These representations diversify and enrich the input conditions to the diffusion models, enabling more diverse outputs. Our experimental results demonstrate that Kaleido effectively broadens the diversity of the generated image samples from a given textual description while maintaining high image quality. Furthermore, we show that Kaleido adheres closely to the guidance provided by the generated latent variables, demonstrating its capability to effectively control and direct the image generation process.

arXiv:2405.21022

You Only Scan Once: Efficient Multi-dimension Sequential Modeling with LightNet

Linear attention mechanisms have gained prominence in causal language models due to their linear computational complexity and enhanced speed. However, the inherent decay mechanism in linear attention presents challenges when applied to multi-dimensional sequence modeling tasks, such as image processing and multi-modal learning. In these scenarios, the utilization of sequential scanning to establish a global receptive field necessitates multiple scans for multi-dimensional data, thereby leading to inefficiencies. This paper identifies the inefficiency caused by a multiplicative linear recurrence and proposes an efficient alternative additive linear recurrence to avoid the issue, as it can handle multi-dimensional data within a single scan. We further develop an efficient multi-dimensional sequential modeling framework called LightNet based on the new recurrence. Moreover, we present two new multi-dimensional linear relative positional encoding methods, MD-TPE and MD-LRPE to enhance the model's ability to discern positional information in multi-dimensional scenarios. Our empirical evaluations across various tasks, including image classification, image generation, bidirectional language modeling, and autoregressive language modeling, demonstrate the efficacy of LightNet, showcasing its potential as a versatile and efficient solution for multi-dimensional sequential modeling.

arXiv:2405.20759

Information Theoretic Text-to-Image Alignment

Diffusion models for Text-to-Image (T2I) conditional generation have recently achieved tremendous success. Yet, aligning these models with user's intentions still involves a laborious trial-and-error process, and this challenging alignment problem has attracted considerable attention from the research community. In this work, instead of relying on fine-grained linguistic analyses of prompts, human annotation, or auxiliary vision-language models, we use Mutual Information (MI) to guide model alignment. In brief, our method uses self-supervised fine-tuning and relies on a point-wise (MI) estimation between prompts and images to create a synthetic fine-tuning set for improving model alignment. Our analysis indicates that our method is superior to the state-of-the-art, yet it only requires the pre-trained denoising network of the T2I model itself to estimate MI, and a simple fine-tuning strategy that improves alignment while maintaining image quality.

arXiv:2405.20750공식 코드 · Papers with Code

Diffusion Models Are Innate One-Step Generators

Diffusion Models (DMs) have achieved great success in image generation and other fields. By fine sampling through the trajectory defined by the SDE/ODE solver based on a well-trained score model, DMs can generate remarkable high-quality results. However, this precise sampling often requires multiple steps and is computationally demanding. To address this problem, instance-based distillation methods have been proposed to distill a one-step generator from a DM by having a simpler student model mimic a more complex teacher model. Yet, our research reveals an inherent limitations in these methods: the teacher model, with more steps and more parameters, occupies different local minima compared to the student model, leading to suboptimal performance when the student model attempts to replicate the teacher. To avoid this problem, we introduce a novel distributional distillation method, which uses an exclusive distributional loss. This method exceeds state-of-the-art (SOTA) results while requiring significantly fewer training images. Additionally, we show that DMs' layers are differentially activated at different time steps, leading to an inherent capability to generate images in a single step. Freezing most of the convolutional layers in a DM during distributional distillation enables this innate capability and leads to further performance improvements. Our method achieves the SOTA results on CIFAR-10 (FID 1.54), AFHQv2 64x64 (FID 1.23), FFHQ 64x64 (FID 0.85) and ImageNet 64x64 (FID 1.16) with great efficiency. Most of those results are obtained with only 5 million training images within 6 hours on 8 A100 GPUs.

arXiv:2405.20717공식 코드 · Papers with Code

Cyclic image generation using chaotic dynamics

Successive image generation using cyclic transformations is demonstrated by extending the CycleGAN model to transform images among three different categories. Repeated application of the trained generators produces sequences of images that transition among the different categories. The generated image sequences occupy a more limited region of the image space compared with the original training dataset. Quantitative evaluation using precision and recall metrics indicates that the generated images have high quality but reduced diversity relative to the training dataset. Such successive generation processes are characterized as chaotic dynamics in terms of dynamical system theory. Positive Lyapunov exponents estimated from the generated trajectories confirm the presence of chaotic dynamics, with the Lyapunov dimension of the attractor found to be comparable to the intrinsic dimension of the training data manifold. The results suggest that chaotic dynamics in the image space defined by the deep generative model contribute to the diversity of the generated images, constituting a novel approach for multi-class image generation. This model can be interpreted as an extension of classical associative memory to perform hetero-association among image categories.

arXiv:2405.20685

Enhancing Counterfactual Image Generation Using Mahalanobis Distance with Distribution Preferences in Feature Space

In the realm of Artificial Intelligence (AI), the importance of Explainable Artificial Intelligence (XAI) is increasingly recognized, particularly as AI models become more integral to our lives. One notable single-instance XAI approach is counterfactual explanation, which aids users in comprehending a model's decisions and offers guidance on altering these decisions. Specifically in the context of image classification models, effective image counterfactual explanations can significantly enhance user understanding. This paper introduces a novel method for computing feature importance within the feature space of a black-box model. By employing information fusion techniques, our method maximizes the use of data to address feature counterfactual explanations in the feature space. Subsequently, we utilize an image generation model to transform these feature counterfactual explanations into image counterfactual explanations. Our experiments demonstrate that the counterfactual explanations generated by our method closely resemble the original images in both pixel and feature spaces. Additionally, our method outperforms established baselines, achieving impressive experimental results.

arXiv:2405.20669공식 코드 · Papers with Code

Hybrid Fourier Score Distillation for Efficient One Image to 3D Object Generation

Single image-to-3D generation is pivotal for crafting controllable 3D assets. Given its under-constrained nature, we attempt to leverage 3D geometric priors from a novel view diffusion model and 2D appearance priors from an image generation model to guide the optimization process. We note that there is a disparity between the generation priors of these two diffusion models, leading to their different appearance outputs. Specifically, image generation models tend to deliver more detailed visuals, whereas novel view models produce consistent yet over-smooth results across different views. Directly combining them leads to suboptimal effects due to their appearance conflicts. Hence, we propose a 2D-3D hybrid Fourier Score Distillation objective function, hy-FSD. It optimizes 3D Gaussians using 3D priors in spatial domain to ensure geometric consistency, while exploiting 2D priors in the frequency domain through Fourier transform for better visual quality. hy-FSD can be integrated into existing 3D generation methods and produce significant performance gains. With this technique, we further develop an image-to-3D generation pipeline to create high-quality 3D objects within one minute, named Fourier123. Extensive experiments demonstrate that Fourier123 excels in efficient generation with rapid convergence speed and visually-friendly generation results.