Chaojie Mao

Zeyinzi Jiang, Chaojie Mao, Ziyuan Huang et al.

Parameter-efficient transfer learning (PETL) based on large-scale pre-trained foundation models has achieved great success in various downstream applications. Existing tuning methods, such as prompt, prefix, and adapter, perform task-specific lightweight adjustments to different parts of the original architecture. However, they take effect on only some parts of the pre-trained models, i.e., only the feed-forward layers or the self-attention layers, which leaves the remaining frozen structures unable to adapt to the data distributions of downstream tasks. Further, the existing structures are strongly coupled with the Transformers, hindering parameter-efficient deployment as well as the design flexibility for new approaches. In this paper, we revisit the design paradigm of PETL and derive a unified framework U-Tuning for parameter-efficient transfer learning, which is composed of an operation with frozen parameters and a unified tuner that adapts the operation for downstream applications. The U-Tuning framework can simultaneously encompass existing methods and derive new approaches for parameter-efficient transfer learning, which prove to achieve on-par or better performances on CIFAR-100 and FGVC datasets when compared with existing PETL methods.

Team Wan, Ang Wang, Baole Ai et al.

This report presents Wan, a comprehensive and open suite of video foundation models designed to push the boundaries of video generation. Built upon the mainstream diffusion transformer paradigm, Wan achieves significant advancements in generative capabilities through a series of innovations, including our novel VAE, scalable pre-training strategies, large-scale data curation, and automated evaluation metrics. These contributions collectively enhance the model's performance and versatility. Specifically, Wan is characterized by four key features: Leading Performance: The 14B model of Wan, trained on a vast dataset comprising billions of images and videos, demonstrates the scaling laws of video generation with respect to both data and model size. It consistently outperforms the existing open-source models as well as state-of-the-art commercial solutions across multiple internal and external benchmarks, demonstrating a clear and significant performance superiority. Comprehensiveness: Wan offers two capable models, i.e., 1.3B and 14B parameters, for efficiency and effectiveness respectively. It also covers multiple downstream applications, including image-to-video, instruction-guided video editing, and personal video generation, encompassing up to eight tasks. Consumer-Grade Efficiency: The 1.3B model demonstrates exceptional resource efficiency, requiring only 8.19 GB VRAM, making it compatible with a wide range of consumer-grade GPUs. Openness: We open-source the entire series of Wan, including source code and all models, with the goal of fostering the growth of the video generation community. This openness seeks to significantly expand the creative possibilities of video production in the industry and provide academia with high-quality video foundation models. All the code and models are available at https://github.com/Wan-Video/Wan2.1.

Zhen Han, Zeyinzi Jiang, Yulin Pan et al.

Diffusion models have emerged as a powerful generative technology and have been found to be applicable in various scenarios. Most existing foundational diffusion models are primarily designed for text-guided visual generation and do not support multi-modal conditions, which are essential for many visual editing tasks. This limitation prevents these foundational diffusion models from serving as a unified model in the field of visual generation, like GPT-4 in the natural language processing field. In this work, we propose ACE, an All-round Creator and Editor, which achieves comparable performance compared to those expert models in a wide range of visual generation tasks. To achieve this goal, we first introduce a unified condition format termed Long-context Condition Unit (LCU), and propose a novel Transformer-based diffusion model that uses LCU as input, aiming for joint training across various generation and editing tasks. Furthermore, we propose an efficient data collection approach to address the issue of the absence of available training data. It involves acquiring pairwise images with synthesis-based or clustering-based pipelines and supplying these pairs with accurate textual instructions by leveraging a fine-tuned multi-modal large language model. To comprehensively evaluate the performance of our model, we establish a benchmark of manually annotated pairs data across a variety of visual generation tasks. The extensive experimental results demonstrate the superiority of our model in visual generation fields. Thanks to the all-in-one capabilities of our model, we can easily build a multi-modal chat system that responds to any interactive request for image creation using a single model to serve as the backend, avoiding the cumbersome pipeline typically employed in visual agents. Code and models will be available on the project page: https://ali-vilab.github.io/ace-page/.

Akide Liu, Jinbo Xing, Chaojie Mao et al.

Minute-scale cinematic video generation is a central challenge for generative video models. Existing paradigms address only fragments of this challenge: single-shot extrapolation preserves an anchor but lacks cinematic structure, while multi-shot storytelling imposes structure yet remains free to invent its visual states rather than continue an observed one. We define Multi-Shot Video Extrapolation (MSVE), a task that extends an observed frame or clip into a sequence of cinematically structured shots while preserving anchor state and advancing narrative intent. This setting operates under the finite per-call generation budget of short-video models. We identify three coupled bottlenecks: (1) global planners over-specify unsupported details from full screenplays; (2) shot-level prompts dilute task-relevant state when carrying the complete story; and (3) temporal chaining turns generated frames into a lossy memory in which identity, scene, object, and action state decay. MSVE reveals that long-video failure is not merely a limitation of context length, but a failure of context allocation. We propose Recursive Context Allocation (ReCA), an inference-time framework that allocates context hierarchically across planning and generation. ReCA recursively decomposes MSVE into context-bounded subproblems, invokes frozen generators at leaf nodes, and propagates structured state updates across time. To evaluate this setting, we further propose MSVE-Bench and NB-Q, a source-grounded protocol with prompts purpose-built for 3 to 5 minute long-video generation, a regime not addressed by existing short-clip benchmarks. Compared to previous methods, ReCA improves average normalized score by 8 to 16 percent over the strongest competing controller and improves multi-shot consistency metrics by 28 to 43 percent. View the project page at https://reca.vmv.re.

Chaojie Mao, Chen-Wei Xie, Chongyang Zhong et al.

We present Wan-Image, a unified visual generation system explicitly engineered to paradigm-shift image generation models from casual synthesizers into professional-grade productivity tools. While contemporary diffusion models excel at aesthetic generation, they frequently encounter critical bottlenecks in rigorous design workflows that demand absolute controllability, complex typography rendering, and strict identity preservation. To address these challenges, Wan-Image features a natively unified multi-modal architecture by synergizing the cognitive capabilities of large language models with the high-fidelity pixel synthesis of diffusion transformers, which seamlessly translates highly nuanced user intents into precise visual outputs. It is fundamentally powered by large-scale multi-modal data scaling, a systematic fine-grained annotation engine, and curated reinforcement learning data to surpass basic instruction following and unlock expert-level professional capabilities. These include ultra-long complex text rendering, hyper-diverse portrait generation, palette-guided generation, multi-subject identity preservation, coherent sequential visual generation, precise multi-modal interactive editing, native alpha-channel generation, and high-efficiency 4K synthesis. Across diverse human evaluations, Wan-Image exceeds Seedream 5.0 Lite and GPT Image 1.5 in overall performance, reaching parity with Nano Banana Pro in challenging tasks. Ultimately, Wan-Image revolutionizes visual content creation across e-commerce, entertainment, education, and personal productivity, redefining the boundaries of professional visual synthesis.

Zeyinzi Jiang, Chaojie Mao, Ziyuan Huang et al.

Parameter-efficient tuning has become a trend in transferring large-scale foundation models to downstream applications. Existing methods typically embed some light-weight tuners into the backbone, where both the design and the learning of the tuners are highly dependent on the base model. This work offers a new tuning paradigm, dubbed Res-Tuning, which intentionally unbinds tuners from the backbone. With both theoretical and empirical evidence, we show that popular tuning approaches have their equivalent counterparts under our unbinding formulation, and hence can be integrated into our framework effortlessly. Thanks to the structural disentanglement, we manage to free the design of tuners from the network architecture, facilitating flexible combination of various tuning strategies. We further propose a memory-efficient variant of Res-Tuning, where the bypass i.e., formed by a sequence of tuners) is effectively detached from the main branch, such that the gradients are back-propagated only to the tuners but not to the backbone. Such a detachment also allows one-time backbone forward for multi-task inference. Extensive experiments on both discriminative and generative tasks demonstrate the superiority of our method over existing alternatives from the perspectives of efficacy and efficiency. Project page: $\href{https://res-tuning.github.io/}{\textit{https://res-tuning.github.io/}}$.

Jinbo Xing, Zeyinzi Jiang, Yuxiang Tuo et al.

Recent unified models have made unprecedented progress in both understanding and generation. However, while most of them accept multi-modal inputs, they typically produce only single-modality outputs. This challenge of producing interleaved content is mainly due to training data scarcity and the difficulty of modeling long-range cross-modal context. To address this issue, we decompose interleaved generation into textual planning and visual consistency modeling, and introduce a framework consisting of a planner and a visualizer. The planner produces dense textual descriptions for visual content, while the visualizer synthesizes images accordingly. Under this guidance, we construct large-scale textual-proxy interleaved data (where visual content is represented in text) to train the planner, and curate reference-guided image data to train the visualizer. These designs give rise to Wan-Weaver, which exhibits emergent interleaved generation ability with long-range textual coherence and visual consistency. Meanwhile, the integration of diverse understanding and generation data into planner training enables Wan-Weaver to achieve robust task reasoning and generation proficiency. To assess the model's capability in interleaved generation, we further construct a benchmark that spans a wide range of use cases across multiple dimensions. Extensive experiments demonstrate that, even without access to any real interleaved data, Wan-Weaver achieves superior performance over existing methods.

Haoyou Deng, Keyu Yan, Chaojie Mao et al.

Recent GRPO-based approaches built on flow matching models have shown remarkable improvements in human preference alignment for text-to-image generation. Nevertheless, they still suffer from the sparse reward problem: the terminal reward of the entire denoising trajectory is applied to all intermediate steps, resulting in a mismatch between the global feedback signals and the exact fine-grained contributions at intermediate denoising steps. To address this issue, we introduce \textbf{DenseGRPO}, a novel framework that aligns human preference with dense rewards, which evaluates the fine-grained contribution of each denoising step. Specifically, our approach includes two key components: (1) we propose to predict the step-wise reward gain as dense reward of each denoising step, which applies a reward model on the intermediate clean images via an ODE-based approach. This manner ensures an alignment between feedback signals and the contributions of individual steps, facilitating effective training; and (2) based on the estimated dense rewards, a mismatch drawback between the uniform exploration setting and the time-varying noise intensity in existing GRPO-based methods is revealed, leading to an inappropriate exploration space. Thus, we propose a reward-aware scheme to calibrate the exploration space by adaptively adjusting a timestep-specific stochasticity injection in the SDE sampler, ensuring a suitable exploration space at all timesteps. Extensive experiments on multiple standard benchmarks demonstrate the effectiveness of the proposed DenseGRPO and highlight the critical role of the valid dense rewards in flow matching model alignment.

Shihao Zhao, Yitong Chen, Zeyinzi Jiang et al.

Unified understanding and generation is a highly appealing research direction in multimodal learning. There exist two approaches: one trains a transformer via an auto-regressive paradigm, and the other adopts a two-stage scheme connecting pre-trained understanding and generative models for alignment fine-tuning. The former demands massive data and computing resources unaffordable for ordinary researchers. Though the latter requires a lower training cost, existing works often suffer from limited task coverage or poor generation quality. Both approaches lack the ability to parse input meta-information (such as task type, image resolution, video duration, etc.) and require manual parameter configuration that is tedious and non-intelligent. In this paper, we propose Unison which adopts the two-stage scheme while preserving the capabilities of the pre-trained models well. With an extremely low training cost, we cover a variety of multimodal understanding tasks, including text, image, and video understanding, as well as diverse generation tasks, such as text-to-visual content generation, editing, controllable generation, and IP-based reference generation. We also equip our model with the ability to automatically parse user intentions, determine the target task type, and accurately extract the meta-information required for the corresponding task. This enables full automation of various multimodal tasks without human intervention. Experiments demonstrate that, under a low-cost setting of only 500k training samples and 50 GPU hours, our model can accurately and automatically identify tasks and extract relevant parameters, and achieve superior performance across a variety of understanding and generation tasks.

Yulin Pan, Xiangteng He, Chaojie Mao et al.

Image generation has witnessed significant advancements in the past few years. However, evaluating the performance of image generation models remains a formidable challenge. In this paper, we propose ICE-Bench, a unified and comprehensive benchmark designed to rigorously assess image generation models. Its comprehensiveness could be summarized in the following key features: (1) Coarse-to-Fine Tasks: We systematically deconstruct image generation into four task categories: No-ref/Ref Image Creating/Editing, based on the presence or absence of source images and reference images. And further decompose them into 31 fine-grained tasks covering a broad spectrum of image generation requirements, culminating in a comprehensive benchmark. (2) Multi-dimensional Metrics: The evaluation framework assesses image generation capabilities across 6 dimensions: aesthetic quality, imaging quality, prompt following, source consistency, reference consistency, and controllability. 11 metrics are introduced to support the multi-dimensional evaluation. Notably, we introduce VLLM-QA, an innovative metric designed to assess the success of image editing by leveraging large models. (3) Hybrid Data: The data comes from real scenes and virtual generation, which effectively improves data diversity and alleviates the bias problem in model evaluation. Through ICE-Bench, we conduct a thorough analysis of existing generation models, revealing both the challenging nature of our benchmark and the gap between current model capabilities and real-world generation requirements. To foster further advancements in the field, we will open-source ICE-Bench, including its dataset, evaluation code, and models, thereby providing a valuable resource for the research community.

Zeyinzi Jiang, Zhen Han, Chaojie Mao et al.

Diffusion Transformer has demonstrated powerful capability and scalability in generating high-quality images and videos. Further pursuing the unification of generation and editing tasks has yielded significant progress in the domain of image content creation. However, due to the intrinsic demands for consistency across both temporal and spatial dynamics, achieving a unified approach for video synthesis remains challenging. We introduce VACE, which enables users to perform Video tasks within an All-in-one framework for Creation and Editing. These tasks include reference-to-video generation, video-to-video editing, and masked video-to-video editing. Specifically, we effectively integrate the requirements of various tasks by organizing video task inputs, such as editing, reference, and masking, into a unified interface referred to as the Video Condition Unit (VCU). Furthermore, by utilizing a Context Adapter structure, we inject different task concepts into the model using formalized representations of temporal and spatial dimensions, allowing it to handle arbitrary video synthesis tasks flexibly. Extensive experiments demonstrate that the unified model of VACE achieves performance on par with task-specific models across various subtasks. Simultaneously, it enables diverse applications through versatile task combinations. Project page: https://ali-vilab.github.io/VACE-Page/.

Chaojie Mao, Jingfeng Zhang, Yulin Pan et al.

We report ACE++, an instruction-based diffusion framework that tackles various image generation and editing tasks. Inspired by the input format for the inpainting task proposed by FLUX.1-Fill-dev, we improve the Long-context Condition Unit (LCU) introduced in ACE and extend this input paradigm to any editing and generation tasks. To take full advantage of image generative priors, we develop a two-stage training scheme to minimize the efforts of finetuning powerful text-to-image diffusion models like FLUX.1-dev. In the first stage, we pre-train the model using task data with the 0-ref tasks from the text-to-image model. There are many models in the community based on the post-training of text-to-image foundational models that meet this training paradigm of the first stage. For example, FLUX.1-Fill-dev deals primarily with painting tasks and can be used as an initialization to accelerate the training process. In the second stage, we finetune the above model to support the general instructions using all tasks defined in ACE. To promote the widespread application of ACE++ in different scenarios, we provide a comprehensive set of models that cover both full finetuning and lightweight finetuning, while considering general applicability and applicability in vertical scenarios. The qualitative analysis showcases the superiority of ACE++ in terms of generating image quality and prompt following ability. Code and models will be available on the project page: https://ali-vilab. github.io/ACE_plus_page/.

Zeyinzi Jiang, Chaojie Mao, Yulin Pan et al.

Image diffusion models have been utilized in various tasks, such as text-to-image generation and controllable image synthesis. Recent research has introduced tuning methods that make subtle adjustments to the original models, yielding promising results in specific adaptations of foundational generative diffusion models. Rather than modifying the main backbone of the diffusion model, we delve into the role of skip connection in U-Net and reveal that hierarchical features aggregating long-distance information across encoder and decoder make a significant impact on the content and quality of image generation. Based on the observation, we propose an efficient generative tuning framework, dubbed SCEdit, which integrates and edits Skip Connection using a lightweight tuning module named SC-Tuner. Furthermore, the proposed framework allows for straightforward extension to controllable image synthesis by injecting different conditions with Controllable SC-Tuner, simplifying and unifying the network design for multi-condition inputs. Our SCEdit substantially reduces training parameters, memory usage, and computational expense due to its lightweight tuners, with backward propagation only passing to the decoder blocks. Extensive experiments conducted on text-to-image generation and controllable image synthesis tasks demonstrate the superiority of our method in terms of efficiency and performance. Project page: \url{https://scedit.github.io/}

Zhen Han, Chaojie Mao, Zeyinzi Jiang et al.

Given an original image, image editing aims to generate an image that align with the provided instruction. The challenges are to accept multimodal inputs as instructions and a scarcity of high-quality training data, including crucial triplets of source/target image pairs and multimodal (text and image) instructions. In this paper, we focus on image style editing and present StyleBooth, a method that proposes a comprehensive framework for image editing and a feasible strategy for building a high-quality style editing dataset. We integrate encoded textual instruction and image exemplar as a unified condition for diffusion model, enabling the editing of original image following multimodal instructions. Furthermore, by iterative style-destyle tuning and editing and usability filtering, the StyleBooth dataset provides content-consistent stylized/plain image pairs in various categories of styles. To show the flexibility of StyleBooth, we conduct experiments on diverse tasks, such as text-based style editing, exemplar-based style editing and compositional style editing. The results demonstrate that the quality and variety of training data significantly enhance the ability to preserve content and improve the overall quality of generated images in editing tasks. Project page can be found at https://ali-vilab.github.io/stylebooth-page/.

Yulin Pan, Chaojie Mao, Zeyinzi Jiang et al.

Prior studies have made significant progress in image inpainting guided by either text description or subject image. However, the research on inpainting with flexible guidance or control, i.e., text-only, image-only, and their combination, is still in the early stage. Therefore, in this paper, we introduce the multimodal promptable image inpainting project: a new task model, and data for taming customized image inpainting. We propose LAR-Gen, a novel approach for image inpainting that enables seamless inpainting of specific region in images corresponding to the mask prompt, incorporating both the text prompt and image prompt. Our LAR-Gen adopts a coarse-to-fine manner to ensure the context consistency of source image, subject identity consistency, local semantic consistency to the text description, and smoothness consistency. It consists of three mechanisms: (i) Locate mechanism: concatenating the noise with masked scene image to achieve precise regional editing, (ii) Assign mechanism: employing decoupled cross-attention mechanism to accommodate multi-modal guidance, and (iii) Refine mechanism: using a novel RefineNet to supplement subject details. Additionally, to address the issue of scarce training data, we introduce a novel data engine to automatically extract substantial pairs of data consisting of local text prompts and corresponding visual instances from a vast image data, leveraging publicly available pre-trained large models. Extensive experiments and various application scenarios demonstrate the superiority of LAR-Gen in terms of both identity preservation and text semantic consistency.

Chaojie Mao, Zeyinzi Jiang

Res-Tuning introduces a flexible and efficient paradigm for model tuning, showing that tuners decoupled from the backbone network can achieve performance comparable to traditional methods. Existing methods commonly construct the tuner as a set of trainable low-rank decomposition matrices, positing that a low-rank subspace suffices for adapting pre-trained foundational models to new scenarios. In this work, we present an advanced, efficient tuner augmented with low-rank attention, termed Res-Attn , which also adheres to the Res-Tuning framework. Res-Attn utilizes a parallel multi-head attention module equipped with low-rank projections for query, key, and value to execute streamlined attention operations. Through training this lightweight attention module, Res-Attn facilitates adaptation to new scenarios. Our extensive experiments across a range of discriminative and generative tasks showcase the superior performance of our method when compared to existing alternatives

Zhi-Fan Wu, Tong Wei, Jianwen Jiang et al.

The existence of noisy data is prevalent in both the training and testing phases of machine learning systems, which inevitably leads to the degradation of model performance. There have been plenty of works concentrated on learning with in-distribution (IND) noisy labels in the last decade, i.e., some training samples are assigned incorrect labels that do not correspond to their true classes. Nonetheless, in real application scenarios, it is necessary to consider the influence of out-of-distribution (OOD) samples, i.e., samples that do not belong to any known classes, which has not been sufficiently explored yet. To remedy this, we study a new problem setup, namely Learning with Open-world Noisy Data (LOND). The goal of LOND is to simultaneously learn a classifier and an OOD detector from datasets with mixed IND and OOD noise. In this paper, we propose a new graph-based framework, namely Noisy Graph Cleaning (NGC), which collects clean samples by leveraging geometric structure of data and model predictive confidence. Without any additional training effort, NGC can detect and reject the OOD samples based on the learned class prototypes directly in testing phase. We conduct experiments on multiple benchmarks with different types of noise and the results demonstrate the superior performance of our method against state of the arts.

Chaojie Mao, Yingming Li, Yaqing Zhang et al.

In this work, we present a Multi-Channel deep convolutional Pyramid Person Matching Network (MC-PPMN) based on the combination of the semantic-components and the color-texture distributions to address the problem of person re-identification. In particular, we learn separate deep representations for semantic-components and color-texture distributions from two person images and then employ pyramid person matching network (PPMN) to obtain correspondence representations. These correspondence representations are fused to perform the re-identification task. Further, the proposed framework is optimized via a unified end-to-end deep learning scheme. Extensive experiments on several benchmark datasets demonstrate the effectiveness of our approach against the state-of-the-art literature, especially on the rank-1 recognition rate.

Chaojie Mao, Yingming Li, Zhongfei Zhang et al.

In this work, we present a deep convolutional pyramid person matching network (PPMN) with specially designed Pyramid Matching Module to address the problem of person re-identification. The architecture takes a pair of RGB images as input, and outputs a similiarity value indicating whether the two input images represent the same person or not. Based on deep convolutional neural networks, our approach first learns the discriminative semantic representation with the semantic-component-aware features for persons and then employs the Pyramid Matching Module to match the common semantic-components of persons, which is robust to the variation of spatial scales and misalignment of locations posed by viewpoint changes. The above two processes are jointly optimized via a unified end-to-end deep learning scheme. Extensive experiments on several benchmark datasets demonstrate the effectiveness of our approach against the state-of-the-art approaches, especially on the rank-1 recognition rate.