Generative Adversarial Networks and Image Synthesis

Neueste Paper

• Know When to Hold 'em: Correct-Token Retention in Uniform-State Diffusion Language Models

  Mojtaba Nafez, James Henderson · 2. Oktober 2026

  Uniform-state diffusion models (USDMs) can revise any token at any denoising step, which lets them correct their own mistakes, a key advantage over masked diffusion. Self-correction, however, requires both revising incorrect tokens and retaining correct ones, and we show that current USDMs lack the …

• Diffusion Editing with Soft Mask: Pixel Level Redo of Image and Video with Adjustable Strength

  Candi Zheng, Yuan Lan · 2. Oktober 2026

  Diffusion models with prompt and reference image-guided editing have seen rapid progress, yet they remain too coarse for pixel-level control. One promising direction is to incorporate a soft mask that specifies spatially varying edit strengths but training such fine-grained control demands expensive…

• BAM! Bayesian Anything Model: a foundation model for generative computational imaging

  Alessio Spagnoletti, Charlesquin Kemajou Mbakam, Jonathan Spence, Andr\'es Almansa, Marcelo Pereyra · 1. Oktober 2026

  Generative models are transforming Bayesian computational imaging, yet the field still lacks physics-aware foundation models. Current practice falls into two camps. Large foundation image models are deployed as plug-and-play priors with zero-shot approximate likelihood guidance, which introduces sig…

• Diffusable Latents from Structure-Agnostic Distillation

  Adrien Ramanana Rahary, Nicolas Dufour, Patrick P\'erez, David Picard · 1. Oktober 2026

  Distilling pretrained foundation models into an autoencoder bottleneck improves latent diffusability, enabling diffusion models to converge faster and reach higher sample quality. Standard distillation aligns the latent at each position to a co-located teacher feature, tying the latent layout to the…

• TexTailor: Texture-Preserving Video Virtual Try-On via Adaptive Garment Conditioning

  Zijing Qin, Jun Zhou, Ruicheng Zhang, Jiaqi Hou, Zunnan Xu, Ronghui Li, Zhenyu Xie, Xiu Li · 1. Oktober 2026

  Video virtual try-on has attracted increasing attention due to its broad potential in digital fashion and intelligent e-commerce. However, existing methods primarily focus on low-resolution settings and still face substantial challenges when extended to high-resolution scenarios. These limitations c…

• DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence

  Xu Huang, Ye Huang, Zijun Liao, Yuwei Niu, Xiaojie Li, Menghan Zhou, De Wen Soh, Xiaotong Li, Daquan Zhou · 1. Oktober 2026

  High-compression tokenizers are essential for scaling latent image generative models. However, aggressive compression creates a fundamental tradeoff between reconstruction fidelity and generation efficiency: high compression image encoder always increases the learning difficulty of diffusion trainin…

• TripleFlow: Training-Free Video Object Removal by Bridging Residual Editing and Native Generation

  Songhe Wang, Lifu Wei, Shuolin Xu, Charles A. Kamhoua, David Miller · 1. Oktober 2026

  Video object removal presents a uniquely difficult editing challenge. Because a removal prompt specifies only what to erase rather than what to generate, the model must infer and reconstruct a highly specific occluded background entirely from the surrounding context. Existing training-free methods s…

• Uncertainty-Aware Consistency Distillation for Few-Step Video Generation

  Lingyu Liu, Yaxiong Wang, Li Zhu, Zhedong Zheng · 1. Oktober 2026

  We study few-step video generation, i.e., distilling a multi-step video generator, which typically requires tens of sampling steps, incurring substantial latency and compute, into a few-step student. Consistency distillation is a common recipe, in which a multi-step teacher provides the consistency …

• EPIC: Epipolar-Consistent 360{\deg} Immersive Stereo Video Generation

  Debabrata Mandal, Dongdong Fu, Jonathon Miller, William Villareal, Xi Peng, Praneeth Chakravarthula · 1. Oktober 2026

  Immersive displays can enable rich and diverse virtual experiences. Manually authoring every possible experience to realize this potential, however, is prohibitively expensive, difficult to scale, and impractical. Generative AI models could remove this bottleneck, but today's models are built for co…

• Curating Synthetic Data for Task-Specific Visual Perception

  Saptarshi Neil Sinha, Paul Julius K\"uhn, Michael Weinmann · 1. Oktober 2026

  Synthetic data are most valuable where general-purpose datasets cannot provide the domain-specific priors a task requires, and where manual annotation is expensive, imprecise, or infeasible. In this article we argue that the central question for specialized vision systems is not how to generate more…

• ExploreNet: Learning Where to Explore in Diffusion GRPO

  Shuyue Stella Li, Xiaochuang Han, Yulia Tsvetkov, Luke Zettlemoyer · 1. Oktober 2026

  Group-relative RL methods such as Flow-GRPO post-train image generators by exploring with isotropic Gaussian noise added at every denoising step. This noise decides which rollouts the model learns from, yet it perturbs every channel and spatial position of the latent equally. In this paper, we inste…

• Masked Swingers: Harnessing Data Augmentation to Advance Autoencoders for Self-Supervised Learning

  Anthony Fuller, Scott C. Lowe, Daniel G. Kyrollos, Graham W. Taylor, Evan Shelhamer, James R. Green · 1. Oktober 2026

  Self-supervised learning (SSL) removes the need for annotations and makes models that are capable across more domains than supervised learning. The autoencoder SSL framework learns by reconstructing its own input after information loss through a bottleneck or noise injection. Masked autoencoders (MA…

• E-MoE: Enhanced Mixture-of-Experts for Non-Factorized Diffusion Language Models

  Arseny Ivanov, Alexander Kolesov, Alexander Korotin, Ivan Oseledets, Mikhail Goncharov · 1. Oktober 2026

  Masked diffusion models (MDMs) generate sequences by progressively unmasking several tokens per denoising step, but their reverse process is typically factorized over positions, limiting sample quality in the few-step regime where diffusion's speed advantage over autoregressive decoding matters most…

• FastGuide: Accelerating Reward Guidance for Diffusion Large Language Models

  Darshan Thaker, Lachlan Ewen MacDonald, Ren\'e Vidal · 1. Oktober 2026

  Gradient-based reward guidance provides a flexible way to use downstream reward models to control masked diffusion language models at inference time. However, its computational cost remains high as each decoding iteration incurs expensive diffusion model forward passes and reward model backpropagati…

• D-Scope: Decomposing and Steering Diffusion Transformers with Sparse Autoencoders

  Xinyue Xu, Jiahao Zhang, Lijie Hu, Peter Hase, Hao Wang · 1. Oktober 2026

  Sparse autoencoders (SAEs) reveal visual structure in diffusion transformers (DiTs), but interpreting a feature does not establish whether it can be used to control generation. We introduce D-Scope (Diffusion Scope), a framework that connects feature interpretation to generation control through shar…

• Steering Fields: Adaptive Vector Fields for Safe Image Generation and Beyond

  Simone Facchiano, Jan Eric Lenssen, Bernt Schiele, Wolfgang Stammer, Fabio Galasso, Jonas Fischer · 1. Oktober 2026

  As state-of-the-art text-to-image flow models achieve near-photorealistic quality, controlling their outputs, e.g., suppressing harmful content while promoting benign alternatives, has become a central challenge. The current steering paradigm consists of adding a global steering vector to selected a…

• Learning Normal Diffusion Dynamics for Backdoor Defense in Text-to-Image Models

  Junjian Li, Xiaolong Liu, Peng Sun, Liantao Wu, Linghan Chen, Yudong Gao, Honglong Chen · 1. Oktober 2026

  Backdoor attacks pose a serious threat to the secure deployment of text-to-image (T2I) diffusion models. Existing defenses typically detect backdoors from specific abnormal patterns in internal representations, which may limit their generalizability with the emergence of increasingly diverse attack …

• PartiCam: Camera Controlled Video Generation with Reward Guidance

  Amine Ouasfi, Runjia Li, Junlin Han, Eric Marchand, Philip H. S. Torr, Adnane Boukhayma · 1. Oktober 2026

  We present PartiCam, a training-free Particle filtering rooted method for improved Camera controlled video generation. Generating videos that follow a precisely specified camera trajectory remains challenging for large video diffusion models. Training-free approaches are backbone-agnostic and avoid …

• CAST: Causal Advantage-Structured Training with Spatially Grounded Compositional Rewards for Diffusion Models

  Shu Yu, Chaochao Lu · 1. Oktober 2026

  Online reinforcement learning has been extended to flow matching for diffusion model (DM) image generation. However, this paradigm faces three limitations: (1) Window selection. Existing methods manually set the stochastic differential equation (SDE) sampling window, i.e., the denoising steps where …

• Distilling Diffusion Score Discrepancy for Efficient Training Data Attribution

  Shixuan Liu, Joan Serr\`a, Kin Wai Cheuk, Jinju Kim, Woosung Choi, Yukara Ikemiya, Wei-Hsiang Liao, Jiaqi W. Ma, Yuki Mitsufuji · 1. Oktober 2026

  Training data attribution for diffusion models aims to identify the training samples that influence a generated instance, but existing methods either require costly per-sample gradient computation or query-specific model optimization. Moreover, most methods attribute changes in a proxy loss rather t…

• ReGain: Restoring Subject Fidelity in Personalization on Synthetic Images

  Shubhang Bhatnagar, Ishan Bhatnagar, Viraj Shah, Narendra Ahuja · 1. Oktober 2026

  Text-to-image diffusion models are personalized to a subject by DreamBooth fine-tuning on a handful of its images. Increasingly, these images come from a diffusion model rather than a camera. We show that fine-tuning on such synthetic images degrades subject fidelity, producing oversaturated color a…

• Persistence Forcing: Exploiting Feature Specialization in Pixel-Space Diffusion

  Chong Wang, Zixuan Fu, Shiqi Huang, Siyuan Yang, Hao Cheng, Bihan Wen · 30. September 2026

  Pixel-space diffusion Transformers (DiTs) directly operate on high-dimensional visual data, yet their hidden representations typically undergo uniform refinement across depth. Natural images, however, are inherently organized at different levels of granularity. Global structure can often be represen…

• Tokens Change, Structure Endures: Spectral Watermarking for Generated Speech

  Kanghwi Lee, Kyeongseok Jeong, Jeongmin Liu · 30. September 2026

  Watermarking is a promising tool for establishing the provenance of AI-generated speech. While many neural audio watermarking methods rely on a separately trained watermark generator, token-level watermarking is a training-free alternative that operates directly during generation. Its main weakness …

• From Scores to Samples: Elastic Forcing for Autoregressive Video Generation

  Chi Zhang, Yueyi Liu, Haoyang Shi, Ruichuan An, Haoyu Li, Yuhang Wu, Sen Cui, Miao Liu · 30. September 2026

  Few-step autoregressive video generation commonly relies on Distribution Matching Distillation (DMD), requiring a bidirectional diffusion teacher and an online fake-score model. We instead learn the rollout distribution directly from reference videos, eliminating both score models during post-traini…

• Flow-JEPA: Robust Latent Dynamics for JEPA World Models via Flow Matching

  Yanchen Huo, Ziying Song, Yadan Luo · 30. September 2026

  Joint-Embedding Predictive Architectures (JEPAs) provide a powerful framework for latent world modeling and planning in a reconstruction-free manner. Although numerous JEPA-based approaches have been proposed to mitigate representation collapse, our experiments on localized, out-of-distribution visu…