Generative Adversarial Networks and Image Synthesis

Últimos artículos

• Posterior Collapse as a Phase Transition in Variational Autoencoders

  Zhen Li, Fan Zhang, Zheng Zhang, Yu Chen · 25 de noviembre de 2025

  We investigate the phenomenon of posterior collapse in variational autoencoders (VAEs) from the perspective of statistical physics, and reveal that it constitutes a phase transition governed jointly by data structure and model hyper-parameters. By analyzing the stability of the trivial solution asso…

• Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models

  Daniel Bogdoll, Johannes Jestram, Jonas Rauch, Christin Scheib, Moritz Wittig, J. Marius Z\"ollner · 25 de noviembre de 2025

  In the foreseeable future, autonomous vehicles will require human assistance in situations they can not resolve on their own. In such scenarios, remote assistance from a human can provide the required input for the vehicle to continue its operation. Typical sensors used in autonomous vehicles includ…

• Breaking the Likelihood-Quality Trade-off in Diffusion Models by Merging Pretrained Experts

  Yasin Esfandiari, Stefan Bauer, Sebastian U. Stich, Andrea Dittadi · 25 de noviembre de 2025

  Diffusion models for image generation often exhibit a trade-off between perceptual sample quality and data likelihood: training objectives emphasizing high-noise denoising steps yield realistic images but poor likelihoods, whereas likelihood-oriented training overweights low-noise steps and harms vi…

• SpectraNet: FFT-assisted Deep Learning Classifier for Deepfake Face Detection

  Nithira Jayarathne, Naveen Basnayake, Keshawa Jayasundara, Pasindu Dodampegama, Praveen Wijesinghe, Hirushika Pelagewatta, Kavishka Abeywardana, Sandushan Ranaweera, Chamira Edussooriya · 25 de noviembre de 2025

  Detecting deepfake images is crucial in combating misinformation. We present a lightweight, generalizable binary classification model based on EfficientNet-B6, fine-tuned with transformation techniques to address severe class imbalances. By leveraging robust preprocessing, oversampling, and optimiza…

• Equivariant Deep Equilibrium Models for Imaging Inverse Problems

  Alexander Mehta, Ruangrawee Kitichotkul, Vivek K Goyal, Juli\'an Tachella · 25 de noviembre de 2025

  Equivariant imaging (EI) enables training signal reconstruction models without requiring ground truth data by leveraging signal symmetries. Deep equilibrium models (DEQs) are a powerful class of neural networks where the output is a fixed point of a learned operator. However, training DEQs with comp…

• Prequential posteriors

  Shreya Sinha-Roy, Richard G. Everitt, Christian P. Robert, Ritabrata Dutta · 25 de noviembre de 2025

  Data assimilation is a fundamental task in updating forecasting models upon observing new data, with applications ranging from weather prediction to online reinforcement learning. Deep generative forecasting models (DGFMs) have shown excellent performance in these areas, but assimilating data into s…

• Flow Map Distillation Without Data

  Shangyuan Tong, Nanye Ma, Saining Xie, Tommi Jaakkola · 25 de noviembre de 2025

  State-of-the-art flow models achieve remarkable quality but require slow, iterative sampling. To accelerate this, flow maps can be distilled from pre-trained teachers, a procedure that conventionally requires sampling from an external dataset. We argue that this data-dependency introduces a fundamen…

• Masked Diffusion Models are Secretly Learned-Order Autoregressive Models

  Prateek Garg, Bhavya Kohli, Sunita Sarawagi · 25 de noviembre de 2025

  Masked Diffusion Models (MDMs) have emerged as one of the most promising paradigms for generative modeling over discrete domains. It is known that MDMs effectively train to decode tokens in a random order, and that this ordering has significant performance implications in practice. This observation …

• MIST: Mutual Information Via Supervised Training

  German Gritsai, Megan Richards, Maxime M\'eloux, Kyunghyun Cho, Maxime Peyrard · 25 de noviembre de 2025

  We propose a fully data-driven approach to designing mutual information (MI) estimators. Since any MI estimator is a function of the observed sample from two random variables, we parameterize this function with a neural network (MIST) and train it end-to-end to predict MI values. Training is perform…

• DiM-TS: Bridge the Gap between Selective State Space Models and Time Series for Generative Modeling

  Zihao Yao, Jiankai Zuo, Yaying Zhang · 25 de noviembre de 2025

  Time series data plays a pivotal role in a wide variety of fields but faces challenges related to privacy concerns. Recently, synthesizing data via diffusion models is viewed as a promising solution. However, existing methods still struggle to capture long-range temporal dependencies and complex cha…

• Bringing Stability to Diffusion: Decomposing and Reducing Variance of Training Masked Diffusion Models

  Mengni Jia, Mengyu Zhou, Yihao Liu, Xiaoxi Jiang, Guanjun Jiang · 25 de noviembre de 2025

  Masked diffusion models (MDMs) are a promising alternative to autoregressive models (ARMs), but they suffer from inherently much higher training variance. High variance leads to noisier gradient estimates and unstable optimization, so even equally strong pretrained MDMs and ARMs that are competitive…

• Learning Straight Flows: Variational Flow Matching for Efficient Generation

  Chenrui Ma, Xi Xiao, Tianyang Wang, Xiao Wang, Yanning Shen · 25 de noviembre de 2025

  Flow Matching has limited ability in achieving one-step generation due to its reliance on learned curved trajectories. Previous studies have attempted to address this limitation by either modifying the coupling distribution to prevent interpolant intersections or introducing consistency and mean-vel…

• AdaTok: Adaptive Token Compression with Object-Aware Representations for Efficient Multimodal LLMs

  Xinliang Zhang, Lei Zhu, Hangzhou He, Shuang Zeng, Ourui Fu, Jiakui Hu, Zhengjian Yao, Yanye Lu · 25 de noviembre de 2025

  Multimodal Large Language Models (MLLMs) have demonstrated substantial value in unified text-image understanding and reasoning, primarily by converting images into sequences of patch-level tokens that align with their architectural paradigm. However, patch-level tokenization leads to a quadratic gro…

• ImAgent: A Unified Multimodal Agent Framework for Test-Time Scalable Image Generation

  Kaishen Wang, Ruibo Chen, Tong Zheng, Heng Huang · 25 de noviembre de 2025

  Recent text-to-image (T2I) models have made remarkable progress in generating visually realistic and semantically coherent images. However, they still suffer from randomness and inconsistency with the given prompts, particularly when textual descriptions are vague or underspecified. Existing approac…

• Scaffold Diffusion: Sparse Multi-Category Voxel Structure Generation with Discrete Diffusion

  Justin Jung · 25 de noviembre de 2025

  Generating realistic sparse multi-category 3D voxel structures is difficult due to the cubic memory scaling of voxel structures and moreover the significant class imbalance caused by sparsity. We introduce Scaffold Diffusion, a generative model designed for sparse multi-category 3D voxel structures.…

• Entropic Time Schedulers for Generative Diffusion Models

  Dejan Stancevic, Florian Handke, Luca Ambrogioni · 25 de noviembre de 2025

  The practical performance of generative diffusion models depends on the appropriate choice of the noise scheduling function, which can also be equivalently expressed as a time reparameterization. In this paper, we present a time scheduler that selects sampling points based on entropy rather than uni…

• Studying Classifier(-Free) Guidance From a Classifier-Centric Perspective

  Xiaoming Zhao, Alexander G. Schwing · 25 de noviembre de 2025

  Classifier-free guidance has become a staple for conditional generation with denoising diffusion models. However, a comprehensive understanding of classifier-free guidance is still missing. In this work, we carry out an empirical study to provide a fresh perspective on classifier-free guidance. Conc…

• In-Video Instructions: Visual Signals as Generative Control

  Gongfan Fang, Xinyin Ma, Xinchao Wang · 25 de noviembre de 2025

  Large-scale video generative models have recently demonstrated strong visual capabilities, enabling the prediction of future frames that adhere to the logical and physical cues in the current observation. In this work, we investigate whether such capabilities can be harnessed for controllable image-…

• DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation

  Zehong Ma, Longhui Wei, Shuai Wang, Shiliang Zhang, Qi Tian · 25 de noviembre de 2025

  Pixel diffusion aims to generate images directly in pixel space in an end-to-end fashion. This approach avoids the limitations of VAE in the two-stage latent diffusion, offering higher model capacity. Existing pixel diffusion models suffer from slow training and inference, as they usually model both…

• Evaluating Dataset Watermarking for Fine-tuning Traceability of Customized Diffusion Models: A Comprehensive Benchmark and Removal Approach

  Xincheng Wang, Hanchi Sun, Wenjun Sun, Kejun Xue, Wangqiu Zhou, Jianbo Zhang, Wei Sun, Dandan Zhu, Xiongkuo Min, Jun Jia, Zhijun Fang · 25 de noviembre de 2025

  Recent fine-tuning techniques for diffusion models enable them to reproduce specific image sets, such as particular faces or artistic styles, but also introduce copyright and security risks. Dataset watermarking has been proposed to ensure traceability by embedding imperceptible watermarks into trai…

• Learning Plug-and-play Memory for Guiding Video Diffusion Models

  Selena Song, Ziming Xu, Zijun Zhang, Kun Zhou, Jiaxian Guo, Lianhui Qin, Biwei Huang · 25 de noviembre de 2025

  Diffusion Transformer(DiT) based video generation models have recently achieved impressive visual quality and temporal coherence, but they still frequently violate basic physical laws and commonsense dynamics, revealing a lack of explicit world knowledge. In this work, we explore how to equip them w…

• From Pixels to Posts: Retrieval-Augmented Fashion Captioning and Hashtag Generation

  Moazzam Umer Gondal, Hamad Ul Qudous, Daniya Siddiqui, Asma Ahmad Farhan · 25 de noviembre de 2025

  This paper introduces the retrieval-augmented framework for automatic fashion caption and hashtag generation, combining multi-garment detection, attribute reasoning, and Large Language Model (LLM) prompting. The system aims to produce visually grounded, descriptive, and stylistically interesting tex…

• EnfoPath: Energy-Informed Analysis of Generative Trajectories in Flow Matching

  Ziyun Li, Ben Dai, Huancheng Hu, Henrik Bostr\"om, Soon Hoe Lim · 25 de noviembre de 2025

  Flow-based generative models synthesize data by integrating a learned velocity field from a reference distribution to the target data distribution. Prior work has focused on endpoint metrics (e.g., fidelity, likelihood, perceptual quality) while overlooking a deeper question: what do the sampling tr…

• Understanding, Accelerating, and Improving MeanFlow Training

  Jin-Young Kim, Hyojun Go, Lea Bogensperger, Julius Erbach, Nikolai Kalischek, Federico Tombari, Konrad Schindler, Dominik Narnhofer · 25 de noviembre de 2025

  MeanFlow promises high-quality generative modeling in few steps, by jointly learning instantaneous and average velocity fields. Yet, the underlying training dynamics remain unclear. We analyze the interaction between the two velocities and find: (i) well-established instantaneous velocity is a prere…

• SpecDiff: Accelerating Diffusion Model Inference with Self-Speculation

  Jiayi Pan, Jiaming Xu, Yongkang Zhou, Guohao Dai · 25 de noviembre de 2025

  Feature caching has recently emerged as a promising method for diffusion model acceleration. It effectively alleviates the inefficiency problem caused by high computational requirements by caching similar features in the inference process of the diffusion model. In this paper, we analyze existing fe…