论文检索

Rong-Cheng Tu, Zi-Ao Ma, Tian Lan, Yuehao Zhao, Heyan Huang, Xian-Ling Mao

Driven by the remarkable progress in diffusion models, text-to-image generation has achieved substantial advancements, underscoring the urgent need for robust automatic quality assessment. This task is inherently complex, requiring evaluations that range from object presence and attribute correctness to relational consistency and visual fidelity. Consequently, current state-of-the-art MLLM-based approaches often rely on powerful commercial models such as GPT-4o, which offer superior reasoning and instruction-following capabilities but are not universally accessible. In contrast, while open-source MLLMs demonstrate promising skills in vision and language understanding, they underperform in comprehensive image quality assessment.To address these challenges, we propose a task decomposition evaluation framework based on GPT-4o to automatically construct a specialized training dataset, breaking down the multifaceted evaluation process into simpler sub-tasks and thus reducing learning complexity. Building on this dataset, we design novel training strategies to distill GPT-4o’s evaluation capabilities into a 7\text{B} open-source MLLM, MiniCPM-V-2.6, enabling it to better follow instructions across diverse assessment criteria. Furthermore, to reliably and comprehensively assess prior works and our proposed model, we manually annotate a meta-evaluation benchmark that includes chain-of-thought explanations alongside quality scores for generated images.Experimental results demonstrate that our distilled open-source MLLM significantly outperforms the current state-of-the-art GPT-4o-base baseline, VIEScore, with over 4.6% improvement in Spearman and Kendall correlations with human judgments.

Jaesung Tae, Hamish Ivison, Sachin Kumar, Arman Cohan

We introduce TESS 2, a general instruction-following diffusion language model that outperforms contemporary instruction-tuned diffusion models, as well as matches and sometimes exceeds strong autoregressive (AR) models. We train TESS 2 by first adapting a strong AR model via continued pretraining with a diffusion loss and then performing further instruction tuning. We find that adaptation training as well as the choice of the base model is crucial for training good instruction-following diffusion models. We further propose reward guidance, a novel and modular inference-time guidance procedure to align model outputs without needing to train the underlying model. Finally, we show that TESS 2 further improves with increased inference-time compute, highlighting the utility of diffusion LMs in having fine-grained controllability over the amount of compute used at inference time.

Hanwen Du, Bo Peng, Xia Ning

Target-Oriented Dialogue (TOD) remains a significant challenge in the LLM era, where strategic dialogue planning is crucial for directing conversations toward specific targets. However, existing dialogue planning methods generate dialogue plans in a step-by-step sequential manner, and may suffer from compounding errors and myopic actions. To address these limitations, we introduce a novel dialogue planning framework, DiffTOD, which leverages diffusion models to enable non-sequential dialogue planning. DiffTOD formulates dialogue planning as a trajectory generation problem with conditional guidance, and leverages a diffusion language model to estimate the likelihood of the dialogue trajectory. To optimize the dialogue action strategies, DiffTOD introduces three tailored guidance mechanisms for different target types, offering flexible guidance toward diverse TOD targets at test time. Extensive experiments across three diverse TOD settings show that DiffTOD can effectively perform non-myopic lookahead exploration and optimize action strategies over a long horizon through non-sequential dialogue planning, and demonstrates strong flexibility across complex and diverse dialogue scenarios. Our code and data are accessible through https://github.com/ninglab/DiffTOD.

Ruizhe Chen, Wenhao Chai, Zhifei Yang, Xiaotian Zhang, Ziyang Wang, Tony Quek, Joey Tianyi Zhou, Soujanya Poria, Zuozhu Liu

Inference-time alignment provides an efficient alternative for aligning LLMs with humans. However, these approaches still face challenges, such as limited scalability due to policy-specific value functions and latency during the inference phase. In this paper, we propose a novel approach, Diffusion-styled Preference Optimization (DiffPO), which provides an efficient and policy-agnostic solution for aligning LLMs with humans. By directly performing alignment at sentence level, DiffPO avoids the time latency associated with token-level generation. Designed as a plug-and-play module, DiffPO can be seamlessly integrated with various base models to enhance their alignment. Extensive experiments on AlpacaEval 2, MT-bench, and HH-RLHF demonstrate that DiffPO achieves superior alignment performance across various settings, achieving a favorable trade-off between alignment quality and inference-time latency. Furthermore, DiffPO demonstrates model-agnostic scalability, significantly improving the performance of large models such as Llama-3-70B.

Jialong Zuo, Shengpeng Ji, Minghui Fang, Mingze Li, Ziyue Jiang, Xize Cheng, Xiaoda Yang, Chen Feiyang, Xinyu Duan, Zhou Zhao

Zero-Shot Voice Conversion (VC) aims to transform the source speaker’s timbre into an arbitrary unseen one while retaining speech content. Most prior work focuses on preserving the source’s prosody, while fine-grained timbre information may leak through prosody, and transferring target prosody to synthesized speech is rarely studied. In light of this, we propose R-VC, a rhythm-controllable and efficient zero-shot voice conversion model. R-VC employs data perturbation techniques and discretize source speech into Hubert content tokens, eliminating much content-irrelevant information. By leveraging a Mask Generative Transformer for in-context duration modeling, our model adapts the linguistic content duration to the desired target speaking style, facilitating the transfer of the target speaker’s rhythm. Furthermore, R-VC introduces a powerful Diffusion Transformer (DiT) with shortcut flow matching during training, conditioning the network not only on the current noise level but also on the desired step size, enabling high timbre similarity and quality speech generation in fewer sampling steps, even in just two, thus minimizing latency. Experimental results show that R-VC achieves comparable speaker similarity to state-of-the-art VC methods with a smaller dataset, and surpasses them in terms of speech naturalness, intelligibility and style transfer performance.

Wenqiang Wang, Yan Xiao, Hao Lin, Yangshijie Zhang, Xiaochun Cao

Current multi-task adversarial text attacks rely on abundant access to shared internal features and numerous queries, often limited to a single task type. As a result, these attacks are less effective against practical scenarios involving black-box feedback APIs, limited queries, or multiple task types. To bridge this gap, we propose Cluster and Ensemble Mutil-task Text Adversarial Attack (CEMA), an effective black-box attack that exploits the transferability of adversarial texts across different tasks. CEMA simplifies complex multi-task scenarios by using a deep-level substitute model trained in a plug-and-play manner for text classification, enabling attacks without mimicking the victim model. This approach requires only a few queries for training, converting multi-task attacks into classification attacks and allowing attacks across various tasks. CEMA generates multiple adversarial candidates using different text classification methods and selects the one that most effectively attacks substitute models. In experiments involving multi-task models with two, three, or six tasks—spanning classification, translation, summarization, and text-to-image generation—CEMA demonstrates significant attack success with as few as 100 queries. Furthermore, CEMA can target commercial APIs (e.g., Baidu and Google Translate), large language models (e.g., ChatGPT 4o), and image-generation models (e.g., Stable Diffusion V2), showcasing its versatility and effectiveness in real-world applications.

Huadai Liu, Jialei Wang, Rongjie Huang, Yang Liu, Heng Lu, Zhou Zhao, Wei Xue

Recent advancements in latent diffusion models (LDMs) have markedly enhanced text-to-audio generation, yet their iterative sampling processes impose substantial computational demands, limiting practical deployment. While recent methods utilizing consistency-based distillation aim to achieve few-step or single-step inference, their one-step performance is constrained by curved trajectories, preventing them from surpassing traditional diffusion models. In this work, we introduce FlashAudio with rectified flows to learn straight flow for fast simulation. To alleviate the inefficient timesteps allocation and suboptimal distribution of noise, FlashAudio optimizes the time distribution of rectified flow with Bifocal Samplers and proposes immiscible flow to minimize the total distance of data-noise pairs in a batch vias assignment. Furthermore, to address the amplified accumulation error caused by the classifier-free guidance (CFG), we propose Anchored Optimization, which refines the guidance scale by anchoring it to a reference trajectory. Experimental results on text-to-audio generation demonstrate that FlashAudio’s one-step generation performance surpasses the diffusion-based models with hundreds of sampling steps on audio quality and enables a sampling speed of 400x faster than real-time on a single NVIDIA 4090Ti GPU. Code will be available at https://github.com/liuhuadai/FlashAudio. Audio Samples are available at https://FlashAudio-TTA.github.io/.

Bocheng Li, Zhujin Gao, Linli Xu

Diffusion models have emerged as a promising approach for text generation, with recent works falling into two main categories: discrete and continuous diffusion models. Discrete diffusion models apply token corruption independently using categorical distributions, allowing for different diffusion progress across tokens but lacking fine-grained control. Continuous diffusion models map tokens to continuous spaces and apply fine-grained noise, but the diffusion progress is uniform across tokens, limiting their ability to capture semantic nuances. To address these limitations, we propose Non-simultaneous Continuous Diffusion Models (NeoDiff), a novel diffusion model that integrates the strengths of both discrete and continuous approaches. NeoDiff introduces a Poisson diffusion process for the forward process, enabling a flexible and fine-grained noising paradigm, and employs a time predictor for the reverse process to adaptively modulate the denoising progress based on token semantics. Furthermore, NeoDiff utilizes an optimized schedule for inference to ensure more precise noise control and improved performance. Our approach unifies the theories of discrete and continuous diffusion models, offering a more principled and effective framework for text generation. Experimental results on several text generation tasks demonstrate NeoDiff’s superior performance compared to baselines of non-autoregressive continuous and discrete diffusion models, iterative-based methods and autoregressive diffusion-based methods. These results highlight NeoDiff’s potential as a powerful tool for generating high-quality text and advancing the field of diffusion-based text generation.

Zhiyu Zhang, Wei Chen, Youfang Lin, Huaiyu Wan

Recent Continual Learning (CL)-based Temporal Knowledge Graph Reasoning (TKGR) methods focus on significantly reducing computational cost and mitigating catastrophic forgetting caused by fine-tuning models with new data. However, existing CL-based TKGR methods still face two key limitations: (1) They usually one-sidedly reorganize individual historical facts, while overlooking the historical context essential for accurately understanding the historical semantics of these facts; (2) They preserve historical knowledge by simply replaying historical facts, while ignoring the potential conflicts between historical and emerging facts. In this paper, we propose a \textbf{D}eep \textbf{G}enerative \textbf{A}daptive \textbf{R}eplay (DGAR) method, which can generate and adaptively replay historical entity distribution representations from the whole historical context. To address the first challenge, historical context prompts as sampling units are built to preserve the whole historical context information. To overcome the second challenge, a pre-trained diffusion model is adopted to generate the historical distribution. During the generation process, the common features between the historical and current distributions are enhanced under the guidance of the TKGR model. In addition, a layer-by-layer adaptive replay mechanism is designed to effectively integrate historical and current distributions. Experimental results demonstrate that DGAR significantly outperforms baselines in reasoning and mitigating forgetting.

Zhenhao Li, Huichi Zhou, Marek Rei, Lucia Specia

Pretrained language models have significantly advanced performance across various natural language processing tasks. However, adversarial attacks continue to pose a critical challenge to system built using these models, as they can be exploited with carefully crafted adversarial texts. Inspired by the ability of diffusion models to predict and reduce noise in computer vision, we propose a novel and flexible adversarial defense method for language classification tasks, DiffuseDef, which incorporates a diffusion layer as a denoiser between the encoder and the classifier. The diffusion layer is trained on top of the existing classifier, ensuring seamless integration with any model in a plug-and-play manner. During inference, the adversarial hidden state is first combined with sampled noise, then denoised iteratively and finally ensembled to produce a robust text representation. By integrating adversarial training, denoising, and ensembling techniques, we show that DiffuseDef improves over existing adversarial defense methods and achieves state-of-the-art performance against common black-box and white-box adversarial attacks.

Fan Zhang, Shulin Tian, Ziqi Huang, Yu Qiao, Ziwei Liu

Recent advancements in visual generative models have enabled high-quality image and video generation, opening diverse applications. However, evaluating these models often demands sampling hundreds or thousands of images or videos, making the process computationally expensive, especially for diffusion-based models with inherently slow sampling. Moreover, existing evaluation methods rely on rigid pipelines that overlook specific user needs and provide numerical results without clear explanations. In contrast, humans can quickly form impressions of a model’s capabilities by observing only a few samples. To mimic this, we propose the Evaluation Agent framework, which employs human-like strategies for efficient, dynamic, multi-round evaluations using only a few samples per round, while offering detailed, user-tailored analyses. It offers four key advantages: 1) efficiency, 2) promptable evaluation tailored to diverse user needs, 3) explainability beyond single numerical scores, and 4) scalability across various models and tools. Experiments show that Evaluation Agent reduces evaluation time to 10% of traditional methods while delivering comparable results. The Evaluation Agent framework is fully open-sourced to advance research in visual generative models and their efficient evaluation.

Yushen Chen, Zhikang Niu, Ziyang Ma, Keqi Deng, Chunhui Wang, JianZhao JianZhao, Kai Yu, Xie Chen

This paper introduces F5-TTS, a fully non-autoregressive text-to-speech system based on flow matching with Diffusion Transformer (DiT). Without requiring complex designs such as duration model, text encoder, and phoneme alignment, the text input is simply padded with filler tokens to the same length as input speech, and then the denoising is performed for speech generation, which was originally proved feasible by E2 TTS. However, the original design of E2 TTS makes it hard to follow due to its slow convergence and low robustness. To address these issues, we first model the input with ConvNeXt to refine the text representation, making it easy to align with the speech. We further propose an inference-time Sway Sampling strategy, which significantly improves our model’s performance and efficiency. This sampling strategy for flow step can be easily applied to existing flow matching based models without retraining. Our design allows faster training and achieves an inference RTF of 0.15, which is greatly improved compared to state-of-the-art diffusion-based TTS models. Trained on a public 100K hours multilingual dataset, our F5-TTS exhibits highly natural and expressive zero-shot ability, seamless code-switching capability, and speed control efficiency. We have released all codes and checkpoints to promote community development, at https://SWivid.github.io/F5-TTS/.

Tong Liu, Zhixin Lai, Jiawen Wang, Gengyuan Zhang, Shuo Chen, Philip Torr, Vera Demberg, Volker Tresp, Jindong Gu

Diffusion models have recently achieved remarkable advancements in terms of image quality and fidelity to textual prompts. Concurrently, the safety of such generative models has become an area of growing concern. This work introduces a novel type of jailbreak, which triggers T2I models to generate the image with visual text, where the image and the text, although considered to be safe in isolation, combine to form unsafe content. To systematically explore this phenomenon, we propose a dataset to evaluate the current diffusion-based text-to-image (T2I) models under such jailbreak. We benchmark nine representative T2I models, including two closed-source commercial models. Experimental results reveal a concerning tendency to produce unsafe content: all tested models suffer from such type of jailbreak, with rates of unsafe generation ranging from around 10% to 70% where DALL·E 3 demonstrates almost the highest unsafety. In real-world scenarios, various filters such as keyword blocklists, customized prompt filters, and NSFW image filters, are commonly employed to mitigate these risks. We evaluate the effectiveness of such filters against our jailbreak and found that, while these filters may be effective for single modality detection, they fail to work against our jailbreak. We also investigate the underlying reason for such jailbreaks, from the perspective of text rendering capability and training data. Our work provides a foundation for further development towards more secure and reliable T2I models.

Xiaochen Zhu, Georgi Karadzhov, Chenxi Whitehouse, Andreas Vlachos

Diffusion models have shown promise in text generation, but often struggle with generating long, coherent, and contextually accurate text. Token-level diffusion doesn’t model word-order dependencies explicitly and operates on short, fixed output windows, while passage-level diffusion struggles with learning robust representations for long-form text. To address these challenges, we propose Segment-Level Diffusion (SLD), a framework that enhances diffusion-based text generation through text segmentation, robust representation training with adversarial and contrastive learning, and improved latent-space guidance. By segmenting long-form outputs into multiple latent representations and decoding them with an autoregressive decoder, SLD simplifies diffusion predictions and improves scalability. Experiments on four datasets demonstrate that, when compared to other diffusion and autoregressive baselines SLD achieves competitive or superior fluency, coherence, and contextual compatibility in automatic and human evaluations.

Reinforcement Learning · Batch/Offline

Yifu Yuan, Zhenrui Zheng, Zibin Dong, Jianye Hao

Multi-objective Reinforcement Learning (MORL) seeks to develop policies that simultaneously optimize multiple conflicting objectives, but it requires extensive online interactions. Offline MORL provides a promising solution by training on pre-collected datasets to generalize to any preference upon deployment. However, real-world offline datasets are often conservatively and narrowly distributed, failing to comprehensively cover preferences, leading to the emergence of out-of-distribution (OOD) preference areas. Existing offline MORL algorithms exhibit poor generalization to OOD preferences, resulting in policies that do not align with preferences. Leveraging the excellent expressive and generalization capabilities of diffusion models, we propose MODULI (Multi-objective Diffusion Planner with Sliding Guidance), which employs a preference-conditioned diffusion model as a planner to generate trajectories that align with various preferences and derive action for decision-making. To achieve accurate generation, MODULI introduces two return normalization methods under diverse preferences for refining guidance. To further enhance generalization to OOD preferences, MODULI proposes a novel sliding guidance mechanism, which involves training an additional slider adapter to capture the direction of preference changes. Incorporating the slider, it transitions from in-distribution (ID) preferences to generating OOD preferences, patching, and extending the incomplete Pareto front. Extensive experiments on the D4MORL benchmark demonstrate that our algorithm outperforms state-of-the-art Offline MORL baselines, exhibiting excellent generalization to OOD preferences.

Deep Learning · Generative Models and Autoencoders

Anji Liu, Xuejie Liu, Dayuan Zhao, Mathias Niepert, Yitao Liang, Guy Van den Broeck

Non-autoregressive (NAR) generative models are valuable because they can handle diverse conditional generation tasks in a more principled way than their autoregressive (AR) counterparts, which are constrained by sequential dependency requirements. Recent advancements in NAR models, such as diffusion language models, have demonstrated superior performance in unconditional generation compared to AR models (e.g., GPTs) of similar sizes. However, such improvements do not always lead to improved conditional generation performance. We show that a key reason for this gap is the difficulty in generalizing to conditional probability queries unseen during training. As a result, strong unconditional generation performance does not guarantee high-quality conditional generation. This paper proposes Tractable Transformers (Tracformer), a Transformer-based generative model that is more robust to different conditional generation tasks. Unlike existing models that rely solely on global contextual features derived from full inputs, Tracformers incorporate a sparse Transformer encoder to capture both local and global contextual information. This information is routed through a decoder for conditional generation. Empirical results demonstrate that Tracformers achieve state-of-the-art conditional generation performance on text modeling compared to recent diffusion and AR model baselines.

Deep Learning · Generative Models and Autoencoders

Hao Chen, Yujin Han, Fangyi Chen, Xiang Li, Yidong Wang, Jindong Wang, Ze Wang, Zicheng Liu, Difan Zou, Bhiksha Raj

Recent advances in latent diffusion models have demonstrated their effectiveness for high-resolution image synthesis. However, the properties of the latent space from tokenizer for better learning and generation of diffusion models remain under-explored. Theoretically and empirically, we find that improved generation quality is closely tied to the latent distributions with better structure, such as the ones with fewer Gaussian Mixture modes and more discriminative features. Motivated by these insights, we propose MAETok, an autoencoder (AE) leveraging mask modeling to learn semantically rich latent space while maintaining reconstruction fidelity. Extensive experiments validate our analysis, demonstrating that the variational form of autoencoders is not necessary, and a discriminative latent space from AE alone enables state-of-the-art performance on ImageNet generation using only 128 tokens. MAETok achieves significant practical improvements, enabling a gFID of 1.69 with 76× faster training and 31× higher inference throughput for 512×512 generation. Our findings show that the structure of the latent space, rather than variational constraints, is crucial for effective diffusion models. Code and trained models will be released.

Applications · Computer Vision

TianChi Xing, Bonan Li, Congying Han, XINMIN QIU, Zicheng Zhang, Tiande Guo

Multi-view Diffusion has greatly advanced the development of 3D content creation by generating multiple images from distinct views, achieving remarkable photorealistic results. However, existing works are still vulnerable to inconsistent 3D geometric structures (commonly known as Janus Problem) and severe artifacts. In this paper, we introduce MIRROR, a versatile plug-and-play method that rectifies such inconsistencies in a training-free manner, enabling the acquisition of high-fidelity, realistic structures without compromising diversity. Our key idea focuses on tracing the motion trajectory of physical points across adjacent viewpoints, enabling rectifications based on neighboring observations of the same region. Technically, MIRROR comprises two core modules: Trajectory Tracking Module (TTM) for pixel-wise trajectory tracking that labels identical points across views, and Feature Rectification Module (FRM) for explicitly adjustment of each pixel embedding on noisy synthesized images by minimizing the distance to corresponding block features in neighboring views, thereby achieving consistent outputs. Extensive evaluations demonstrate that MIRROR can seamlessly integrate with a diverse range of off-the-shelf object-level multi-view diffusion models, significantly enhancing both the consistency and the fidelity in an efficient way.

Deep Learning · Robustness

Haoyang Luo, Linwei Tao, Minjing Dong, Chang Xu

Model calibration seeks to ensure that models produce confidence scores that accurately reflect the true likelihood of their predictions being correct. However, existing calibration approaches are fundamentally tied to datasets of one-hot labels implicitly assuming full certainty in all the annotations. Such datasets are effective for classification but provides insufficient knowledge of uncertainty for model calibration, necessitating the curation of datasets with numerically rich ground-truth confidence values. However, due to the scarcity of uncertain visual examples, such samples are not easily available as real datasets. In this paper, we introduce calibration-aware data augmentation to create synthetic datasets of diverse samples and their ground-truth uncertainty. Specifically, we present **Calibration-aware Semantic Mixing (CSM)**, a novel framework that generates training samples with mixed class characteristics and annotates them with distinct confidence scores via diffusion models. Based on this framework, we propose calibrated reannotation to tackle the misalignment between the annotated confidence score and the mixing ratio during the diffusion reverse process. Besides, we explore the loss functions that better fit the new data representation paradigm. Experimental results demonstrate that CSM achieves superior calibration compared to the state-of-the-art calibration approaches. Our code is [available here](https://github.com/E-Galois/CSM).

Deep Learning · Everything Else

Qingqing Cao, Mahyar Najibi, Sachin Mehta

Pretraining robust vision or multimodal foundation models (e.g., CLIP) relies on large-scale datasets that may be noisy, potentially misaligned, and have long-tail distributions. Previous works have shown promising results in augmenting datasets by generating synthetic samples. However, they only support domain-specific ad hoc use cases (e.g., either image or text only, but not both), and are limited in data diversity due to a lack of fine-grained control over the synthesis process. In this paper, we design a controllable image-text synthesis pipeline, CtrlSynth, for data-efficient and robust multimodal learning. The key idea is to decompose the visual semantics of an image into basic elements, apply user-specified control policies (e.g., remove, add, or replace operations), and recompose them to synthesize images or texts. The decompose and recompose feature in CtrlSynth allows users to control data synthesis in a fine-grained manner by defining customized control policies to manipulate the basic elements. CtrlSynth leverages the capabilities of pretrained foundation models such as large language models or diffusion models to reason and recompose basic elements such that synthetic samples are natural and composed in diverse ways. CtrlSynth is a closed-loop, training-free, and modular framework, making it easy to support different pretrained models. With extensive experiments on 31 datasets spanning different vision and vision-language tasks, we show that CtrlSynth substantially improves zero-shot classification, image-text retrieval, and compositional reasoning performance of CLIP models.