PyTorchKR Papers

최신 논문

총 7,086편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2502.06814

Diffusion Instruction Tuning

We introduce Lavender, a simple supervised fine-tuning (SFT) method that boosts the performance of advanced vision-language models (VLMs) by leveraging state-of-the-art image generation models such as Stable Diffusion. Specifically, Lavender aligns the text-vision attention in the VLM transformer with the equivalent used by Stable Diffusion during SFT, instead of adapting separate encoders. This alignment enriches the model's visual understanding and significantly boosts performance across in- and out-of-distribution tasks. Lavender requires just 0.13 million training examples, 2.5% of typical large-scale SFT datasets, and fine-tunes on standard hardware (8 GPUs) in a single day. It consistently improves state-of-the-art open-source multimodal LLMs (e.g., Llama-3.2-11B, MiniCPM-Llama3-v2.5), achieving up to 30% gains and a 68% boost on challenging out-of-distribution medical QA tasks. By efficiently transferring the visual expertise of image generators with minimal supervision, Lavender offers a scalable solution for more accurate vision-language systems. All code, training data, and models will be shared at https://astrazeneca.github.io/vlm/.

arXiv:2502.07580공식 코드 · Papers with Code

Generative Modeling with Bayesian Sample Inference

We derive a novel generative model from iterative Gaussian posterior inference. By treating the generated sample as an unknown variable, we can formulate the sampling process in the language of Bayesian probability. Our model uses a sequence of prediction and posterior update steps to iteratively narrow down the unknown sample starting from a broad initial belief. In addition to a rigorous theoretical analysis, we establish a connection between our model and diffusion models and show that it includes Bayesian Flow Networks (BFNs) as a special case. In our experiments, we demonstrate that our model improves sample quality on ImageNet32 over both BFNs and the closely related Variational Diffusion Models, while achieving equivalent log-likelihoods on ImageNet32 and CIFAR10. Find our code at https://github.com/martenlienen/bsi.

arXiv:2502.06823공식 코드 · Papers with Code

CTR-Driven Advertising Image Generation with Multimodal Large Language Models

In web data, advertising images are crucial for capturing user attention and improving advertising effectiveness. Most existing methods generate background for products primarily focus on the aesthetic quality, which may fail to achieve satisfactory online performance. To address this limitation, we explore the use of Multimodal Large Language Models (MLLMs) for generating advertising images by optimizing for Click-Through Rate (CTR) as the primary objective. Firstly, we build targeted pre-training tasks, and leverage a large-scale e-commerce multimodal dataset to equip MLLMs with initial capabilities for advertising image generation tasks. To further improve the CTR of generated images, we propose a novel reward model to fine-tune pre-trained MLLMs through Reinforcement Learning (RL), which can jointly utilize multimodal features and accurately reflect user click preferences. Meanwhile, a product-centric preference optimization strategy is developed to ensure that the generated background content aligns with the product characteristics after fine-tuning, enhancing the overall relevance and effectiveness of the advertising images. Extensive experiments have demonstrated that our method achieves state-of-the-art performance in both online and offline metrics. Our code and pre-trained models are publicly available at: https://github.com/Chenguoz/CAIG.

arXiv:2502.05505공식 코드 · Papers with Code

Differentially Private Synthetic Data via APIs 3: Using Simulators Instead of Foundation Model

Differentially private (DP) synthetic data, which closely resembles the original private data while maintaining strong privacy guarantees, has become a key tool for unlocking the value of private data without compromising privacy. Recently, Private Evolution (PE) has emerged as a promising method for generating DP synthetic data. Unlike other training-based approaches, PE only requires access to inference APIs from foundation models, enabling it to harness the power of state-of-the-art (SoTA) models. However, a suitable foundation model for a specific private data domain is not always available. In this paper, we discover that the PE framework is sufficiently general to allow APIs beyond foundation models. In particular, we demonstrate that many SoTA data synthesizers that do not rely on neural networks--such as computer graphics-based image generators, which we refer to as simulators--can be effectively integrated into PE. This insight significantly broadens PE's applicability and unlocks the potential of powerful simulators for DP data synthesis. We explore this approach, named Sim-PE, in the context of image synthesis. Across four diverse simulators, Sim-PE performs well, improving the downstream classification accuracy of PE by up to 3x, reducing FID by up to 80%, and offering much greater efficiency. We also show that simulators and foundation models can be easily leveraged together within PE to achieve further improvements. The code is open-sourced in the Private Evolution Python library: https://github.com/microsoft/DPSDA.

arXiv:2502.03826

FairT2I: Mitigating Social Bias in Text-to-Image Generation via Large Language Model-Assisted Detection and Attribute Rebalancing

The proliferation of Text-to-Image (T2I) models has revolutionized content creation, providing powerful tools for diverse applications ranging from artistic expression to educational material development and marketing. Despite these technological advancements, significant ethical concerns arise from these models' reliance on large-scale datasets that often contain inherent societal biases. These biases are further amplified when AI-generated content is included in training data, potentially reinforcing and perpetuating stereotypes in the generated outputs. In this paper, we introduce FairT2I, a novel framework that harnesses large language models to detect and mitigate social biases in T2I generation. Our framework comprises two key components: (1) an LLM-based bias detection module that identifies potential social biases in generated images based on text prompts, and (2) an attribute rebalancing module that fine-tunes sensitive attributes within the T2I model to mitigate identified biases. Our extensive experiments across various T2I models and datasets show that FairT2I can significantly reduce bias while maintaining high-quality image generation. We conducted both qualitative user studies and quantitative non-parametric analyses in the generated image feature space, building upon the occupational dataset introduced in the Stable Bias study. Our results show that FairT2I successfully mitigates social biases and enhances the diversity of sensitive attributes in generated images. We further demonstrate, using the P2 dataset, that our framework can detect subtle biases that are challenging for human observers to perceive, extending beyond occupation-related prompts. On the basis of these findings, we introduce a new benchmark dataset for evaluating bias in T2I models.

arXiv:2502.03429

On Fairness of Unified Multimodal Large Language Model for Image Generation

Unified multimodal large language models (U-MLLMs) have demonstrated impressive performance in visual understanding and generation in an end-to-end pipeline. Compared with generation-only models (e.g., Stable Diffusion), U-MLLMs may raise new questions about bias in their outputs, which can be affected by their unified capabilities. This gap is particularly concerning given the under-explored risk of propagating harmful stereotypes. In this paper, we benchmark the latest U-MLLMs and find that most exhibit significant demographic biases, such as gender and race bias. To better understand and mitigate this issue, we propose a locate-then-fix strategy, where we audit and show how the individual model component is affected by bias. Our analysis shows that bias originates primarily from the language model. More interestingly, we observe a "partial alignment" phenomenon in U-MLLMs, where understanding bias appears minimal, but generation bias remains substantial. Thus, we propose a novel balanced preference model to balance the demographic distribution with synthetic data. Experiments demonstrate that our approach reduces demographic bias while preserving semantic fidelity. We hope our findings underscore the need for more holistic interpretation and debiasing strategies of U-MLLMs in the future.

arXiv:2502.04116

Generative Adversarial Networks Bridging Art and Machine Intelligence

Generative Adversarial Networks (GAN) have greatly influenced the development of computer vision and artificial intelligence in the past decade and also connected art and machine intelligence together. This book begins with a detailed introduction to the fundamental principles and historical development of GANs, contrasting them with traditional generative models and elucidating the core adversarial mechanisms through illustrative Python examples. The text systematically addresses the mathematical and theoretical underpinnings including probability theory, statistics, and game theory providing a solid framework for understanding the objectives, loss functions, and optimisation challenges inherent to GAN training. Subsequent chapters review classic variants such as Conditional GANs, DCGANs, InfoGAN, and LAPGAN before progressing to advanced training methodologies like Wasserstein GANs, GANs with gradient penalty, least squares GANs, and spectral normalisation techniques. The book further examines architectural enhancements and task-specific adaptations in generators and discriminators, showcasing practical implementations in high resolution image generation, artistic style transfer, video synthesis, text to image generation and other multimedia applications. The concluding sections offer insights into emerging research trends, including self-attention mechanisms, transformer-based generative models, and a comparative analysis with diffusion models, thus charting promising directions for future developments in both academic and applied settings.

arXiv:2502.03426

TruePose: Human-Parsing-guided Attention Diffusion for Full-ID Preserving Pose Transfer

Pose-Guided Person Image Synthesis (PGPIS) generates images that maintain a subject's identity from a source image while adopting a specified target pose (e.g., skeleton). While diffusion-based PGPIS methods effectively preserve facial features during pose transformation, they often struggle to accurately maintain clothing details from the source image throughout the diffusion process. This limitation becomes particularly problematic when there is a substantial difference between the source and target poses, significantly impacting PGPIS applications in the fashion industry where clothing style preservation is crucial for copyright protection. Our analysis reveals that this limitation primarily stems from the conditional diffusion model's attention modules failing to adequately capture and preserve clothing patterns. To address this limitation, we propose human-parsing-guided attention diffusion, a novel approach that effectively preserves both facial and clothing appearance while generating high-quality results. We propose a human-parsing-aware Siamese network that consists of three key components: dual identical UNets (TargetNet for diffusion denoising and SourceNet for source image embedding extraction), a human-parsing-guided fusion attention (HPFA), and a CLIP-guided attention alignment (CAA). The HPFA and CAA modules can embed the face and clothes patterns into the target image generation adaptively and effectively. Extensive experiments on both the in-shop clothes retrieval benchmark and the latest in-the-wild human editing dataset demonstrate our method's significant advantages over 13 baseline approaches for preserving both facial and clothes appearance in the source image.

arXiv:2502.05178

QLIP: Text-Aligned Visual Tokenization Unifies Auto-Regressive Multimodal Understanding and Generation

We introduce Quantized Language-Image Pretraining (QLIP), a visual tokenization method that combines state-of-the-art reconstruction quality with state-of-the-art zero-shot image understanding. QLIP trains a binary-spherical-quantization-based autoencoder with reconstruction and language-image alignment objectives. We are the first to show that the two objectives do not need to be at odds. We balance the two loss terms dynamically during training and show that a two-stage training pipeline effectively mixes the large-batch requirements of image-language pre-training with the memory bottleneck imposed by the reconstruction objective. We validate the effectiveness of QLIP for multimodal understanding and text-conditioned image generation with a single model. Specifically, QLIP serves as a drop-in replacement for the visual encoder for LLaVA and the image tokenizer for LlamaGen with comparable or even better performance. Finally, we demonstrate that QLIP enables a unified mixed-modality auto-regressive model for understanding and generation.

arXiv:2502.05153

Hummingbird: High Fidelity Image Generation via Multimodal Context Alignment

While diffusion models are powerful in generating high-quality, diverse synthetic data for object-centric tasks, existing methods struggle with scene-aware tasks such as Visual Question Answering (VQA) and Human-Object Interaction (HOI) Reasoning, where it is critical to preserve scene attributes in generated images consistent with a multimodal context, i.e. a reference image with accompanying text guidance query. To address this, we introduce Hummingbird, the first diffusion-based image generator which, given a multimodal context, generates highly diverse images w.r.t. the reference image while ensuring high fidelity by accurately preserving scene attributes, such as object interactions and spatial relationships from the text guidance. Hummingbird employs a novel Multimodal Context Evaluator that simultaneously optimizes our formulated Global Semantic and Fine-grained Consistency Rewards to ensure generated images preserve the scene attributes of reference images in relation to the text guidance while maintaining diversity. As the first model to address the task of maintaining both diversity and fidelity given a multimodal context, we introduce a new benchmark formulation incorporating MME Perception and Bongard HOI datasets. Benchmark experiments show Hummingbird outperforms all existing methods by achieving superior fidelity while maintaining diversity, validating Hummingbird's potential as a robust multimodal context-aligned image generator in complex visual tasks.

arXiv:2502.04923공식 코드 · Papers with Code

Cached Multi-Lora Composition for Multi-Concept Image Generation

Low-Rank Adaptation (LoRA) has emerged as a widely adopted technique in text-to-image models, enabling precise rendering of multiple distinct elements, such as characters and styles, in multi-concept image generation. However, current approaches face significant challenges when composing these LoRAs for multi-concept image generation, resulting in diminished generated image quality. In this paper, we initially investigate the role of LoRAs in the denoising process through the lens of the Fourier frequency domain. Based on the hypothesis that applying multiple LoRAs could lead to "semantic conflicts", we find that certain LoRAs amplify high-frequency features such as edges and textures, whereas others mainly focus on low-frequency elements, including the overall structure and smooth color gradients. Building on these insights, we devise a frequency domain based sequencing strategy to determine the optimal order in which LoRAs should be integrated during inference. This strategy offers a methodical and generalizable solution compared to the naive integration commonly found in existing LoRA fusion techniques. To fully leverage our proposed LoRA order sequence determination method in multi-LoRA composition tasks, we introduce a novel, training-free framework, Cached Multi-LoRA (CMLoRA), designed to efficiently integrate multiple LoRAs while maintaining cohesive image generation. With its flexible backbone for multi-LoRA fusion and a non-uniform caching strategy tailored to individual LoRAs, CMLoRA has the potential to reduce semantic conflicts in LoRA composition and improve computational efficiency. Our experimental evaluations demonstrate that CMLoRA outperforms state-of-the-art training-free LoRA fusion methods by a significant margin -- it achieves an average improvement of $2.19\%$ in CLIPScore, and $11.25\%$ in MLLM win rate compared to LoraHub, LoRA Composite, and LoRA Switch.

arXiv:2502.04896

Goku: Flow Based Video Generative Foundation Models

This paper introduces Goku, a state-of-the-art family of joint image-and-video generation models leveraging rectified flow Transformers to achieve industry-leading performance. We detail the foundational elements enabling high-quality visual generation, including the data curation pipeline, model architecture design, flow formulation, and advanced infrastructure for efficient and robust large-scale training. The Goku models demonstrate superior performance in both qualitative and quantitative evaluations, setting new benchmarks across major tasks. Specifically, Goku achieves 0.76 on GenEval and 83.65 on DPG-Bench for text-to-image generation, and 84.85 on VBench for text-to-video tasks. We believe that this work provides valuable insights and practical advancements for the research community in developing joint image-and-video generation models.

arXiv:2502.04412

Decoder-Only LLMs are Better Controllers for Diffusion Models

Groundbreaking advancements in text-to-image generation have recently been achieved with the emergence of diffusion models. These models exhibit a remarkable ability to generate highly artistic and intricately detailed images based on textual prompts. However, obtaining desired generation outcomes often necessitates repetitive trials of manipulating text prompts just like casting spells on a magic mirror, and the reason behind that is the limited capability of semantic understanding inherent in current image generation models. Specifically, existing diffusion models encode the text prompt input with a pre-trained encoder structure, which is usually trained on a limited number of image-caption pairs. The state-of-the-art large language models (LLMs) based on the decoder-only structure have shown a powerful semantic understanding capability as their architectures are more suitable for training on very large-scale unlabeled data. In this work, we propose to enhance text-to-image diffusion models by borrowing the strength of semantic understanding from large language models, and devise a simple yet effective adapter to allow the diffusion models to be compatible with the decoder-only structure. Meanwhile, we also provide a supporting theoretical analysis with various architectures (e.g., encoder-only, encoder-decoder, and decoder-only), and conduct extensive empirical evaluations to verify its effectiveness. The experimental results show that the enhanced models with our adapter module are superior to the stat-of-the-art models in terms of text-to-image generation quality and reliability.

arXiv:2502.04684

G2PDiffusion: Genotype-to-Phenotype Prediction with Diffusion Models

Discovering the genotype-phenotype relationship is crucial for genetic engineering, which will facilitate advances in fields such as crop breeding, conservation biology, and personalized medicine. Current research usually focuses on single species and small datasets due to limitations in phenotypic data collection, especially for traits that require visual assessments or physical measurements. Deciphering complex and composite phenotypes, such as morphology, from genetic data at scale remains an open question. To break through traditional generic models that rely on simplified assumptions, this paper introduces G2PDiffusion, the first-of-its-kind diffusion model designed for genotype-to-phenotype generation across multiple species. Specifically, we use images to represent morphological phenotypes across species and redefine phenotype prediction as conditional image generation. To this end, this paper introduces an environment-enhanced DNA sequence conditioner and trains a stable diffusion model with a novel alignment method to improve genotype-to-phenotype consistency. Extensive experiments demonstrate that our approach enhances phenotype prediction accuracy across species, capturing subtle genetic variations that contribute to observable traits.

arXiv:2502.04475

Augmented Conditioning Is Enough For Effective Training Image Generation

Image generation abilities of text-to-image diffusion models have significantly advanced, yielding highly photo-realistic images from descriptive text and increasing the viability of leveraging synthetic images to train computer vision models. To serve as effective training data, generated images must be highly realistic while also sufficiently diverse within the support of the target data distribution. Yet, state-of-the-art conditional image generation models have been primarily optimized for creative applications, prioritizing image realism and prompt adherence over conditional diversity. In this paper, we investigate how to improve the diversity of generated images with the goal of increasing their effectiveness to train downstream image classification models, without fine-tuning the image generation model. We find that conditioning the generation process on an augmented real image and text prompt produces generations that serve as effective synthetic datasets for downstream training. Conditioning on real training images contextualizes the generation process to produce images that are in-domain with the real image distribution, while data augmentations introduce visual diversity that improves the performance of the downstream classifier. We validate augmentation-conditioning on a total of five established long-tail and few-shot image classification benchmarks and show that leveraging augmentations to condition the generation process results in consistent improvements over the state-of-the-art on the long-tailed benchmark and remarkable gains in extreme few-shot regimes of the remaining four benchmarks. These results constitute an important step towards effectively leveraging synthetic data for downstream training.

arXiv:2502.04391

Towards Fair and Robust Face Parsing for Generative AI: A Multi-Objective Approach

Face parsing is a fundamental task in computer vision, enabling applications such as identity verification, facial editing, and controllable image synthesis. However, existing face parsing models often lack fairness and robustness, leading to biased segmentation across demographic groups and errors under occlusions, noise, and domain shifts. These limitations affect downstream face synthesis, where segmentation biases can degrade generative model outputs. We propose a multi-objective learning framework that optimizes accuracy, fairness, and robustness in face parsing. Our approach introduces a homotopy-based loss function that dynamically adjusts the importance of these objectives during training. To evaluate its impact, we compare multi-objective and single-objective U-Net models in a GAN-based face synthesis pipeline (Pix2PixHD). Our results show that fairness-aware and robust segmentation improves photorealism and consistency in face generation. Additionally, we conduct preliminary experiments using ControlNet, a structured conditioning model for diffusion-based synthesis, to explore how segmentation quality influences guided image generation. Our findings demonstrate that multi-objective face parsing improves demographic consistency and robustness, leading to higher-quality GAN-based synthesis.

arXiv:2502.04991

Bridging Scales in Map Generation: A scale-aware cascaded generative mapping framework for seamless and consistent multi-scale cartographic representation

Multi-scale tile maps are essential for geographic information services, serving as fundamental outcomes of surveying and cartographic workflows. While existing image generation networks can produce map-like outputs from remote sensing imagery, their emphasis on replicating texture rather than preserving geospatial features limits cartographic validity. Current approaches face two fundamental challenges: inadequate integration of cartographic generalization principles with dynamic multi-scale generation and spatial discontinuities arising from tile-wise generation. To address these limitations, we propose a scale-aware cartographic generation framework (SCGM) that leverages conditional guided diffusion and a multi-scale cascade architecture. The framework introduces three key innovations: a scale modality encoding mechanism to formalize map generalization relationships, a scale-driven conditional encoder for robust feature fusion, and a cascade reference mechanism ensuring cross-scale visual consistency. By hierarchically constraining large-scale map synthesis with small-scale structural priors, SCGM effectively mitigates edge artifacts while maintaining geographic fidelity. Comprehensive evaluations on cartographic benchmarks confirm the framework's ability to generate seamless multi-scale tile maps with enhanced spatial coherence and generalization-aware representation, demonstrating significant potential for emergency mapping and automated cartography applications.

arXiv:2502.03079

Poisson Flow Joint Model for Multiphase contrast-enhanced CT

In clinical practice, multiphase contrast-enhanced CT (MCCT) is important for physiological and pathological imaging with contrast injection, which undergoes non-contrast, venous, and delayed phases. Inevitably, the accumulated radiation dose to a patient is higher for multiphase scans than for a plain CT scan. Low-dose CECT is thus highly desirable, but it often leads to suboptimal image quality due to reduced radiation dose. Recently, a generalized Poisson flow generative model (PFGM++) was proposed to unify the diffusion model and the Poisson flow generative models (PFGM), and outperform either of them with an optimized dimensionality of the augmentation data space, holding a significant promise for generic or conditional image generation. In this paper, we propose a Poisson flow joint model (PFJM) for low-dose MCCT to suppress image noise and preserve clinical features. Our model is built on the PFGM++ architecture to transform the multiphase imaging problem into learning the joint distribution of routine-dose MCCT images by optimizing a task-specific generation path with respect to the dimensionality D of the augmented data space. Then, our PFJM model takes the joint low-dose MCCT images as the condition and robustly drives the generative trajectory towards the solution in the routine-dose MCCT domain. Extensive experiments demonstrate that our model is favorably compared with competing models, with MAE of 8.99 HU, SSIM of 98.75% and PSNR of 48.24db, as averaged over all the phases.

arXiv:2502.04308공식 코드 · Papers with Code

HOG-Diff: Higher-Order Guided Diffusion for Graph Generation

Graph generation is a critical yet challenging task as empirical analyses require a deep understanding of complex, non-Euclidean structures. Although diffusion models have recently made significant achievements in graph generation, these models typically adapt from the frameworks designed for image generation, making them ill-suited for capturing the topological properties of graphs. In this work, we propose a novel Higher-order Guided Diffusion (HOG-Diff) model that follows a coarse-to-fine generation curriculum and is guided by higher-order information, enabling the progressive generation of plausible graphs with inherent topological structures. We further prove that our model exhibits a stronger theoretical guarantee than classical diffusion frameworks. Extensive experiments on both molecular and generic graph generation tasks demonstrate that our method consistently outperforms or remains competitive with state-of-the-art baselines. Our code is available at https://github.com/Yiminghh/HOG-Diff.

arXiv:2502.02771

When are Diffusion Priors Helpful in Sparse Reconstruction? A Study with Sparse-view CT

Diffusion models demonstrate state-of-the-art performance on image generation, and are gaining traction for sparse medical image reconstruction tasks. However, compared to classical reconstruction algorithms relying on simple analytical priors, diffusion models have the dangerous property of producing realistic looking results \emph{even when incorrect}, particularly with few observations. We investigate the utility of diffusion models as priors for image reconstruction by varying the number of observations and comparing their performance to classical priors (sparse and Tikhonov regularization) using pixel-based, structural, and downstream metrics. We make comparisons on low-dose chest wall computed tomography (CT) for fat mass quantification. First, we find that classical priors are superior to diffusion priors when the number of projections is ``sufficient''. Second, we find that diffusion priors can capture a large amount of detail with very few observations, significantly outperforming classical priors. However, they fall short of capturing all details, even with many observations. Finally, we find that the performance of diffusion priors plateau after extremely few ($\approx$10-15) projections. Ultimately, our work highlights potential issues with diffusion-based sparse reconstruction and underscores the importance of further investigation, particularly in high-stakes clinical settings.