CVPR 2025 论文列表

Controllable person image generation aims to generate a person image conditioned on reference images, allowing precise control over the person's appearance or pose.However, prior methods often distort fine-grained textural details from the reference image, despite achieving high overall image quality.We attribute these distortions to inadequate attention to corresponding regions in the reference image.To address this, we thereby propose learning flow fields in attention (Leffa), which explicitly guides the target query to attend to the correct reference key in the attention layer during training.Specifically, it is realized via a regularization loss on top of the attention map within a diffusion-based baseline.Our extensive experiments show that Leffa achieves state-of-the-art performance in controlling appearance (virtual try-on) and pose (pose transfer), significantly reducing fine-grained detail distortion while maintaining high image quality.Additionally, we show that our loss is model-agnostic and can be used to improve the performance of other diffusion models.

The latest emerged 4D Panoptic Scene Graph (4D-PSG) provides an advanced-ever representation for comprehensively modeling the dynamic 4D visual real world. Unfortunately, current pioneering 4D-PSG research can largely suffer from data scarcity issues severely, as well as the resulting out-of-vocabulary problems; also, the pipeline nature of the benchmark generation method can lead to suboptimal performance. To address these challenges, this paper investigates a novel framework for 4D-PSG generation that leverages rich 2D visual scene annotations to enhance 4D scene learning. First, we introduce a 4D Large Language Model (4D-LLM) integrated with a 3D mask decoder for end-to-end generation of 4D-PSG. A chained SG inference mechanism is further designed to exploit LLMs' open-vocabulary capabilities to infer accurate and comprehensive object and relation labels iteratively. Most importantly, we propose a 2D-to-4D visual scene transfer learning framework, where a spatial-temporal scene transcending strategy effectively transfers dimension-invariant features from abundant 2D SG annotations to 4D scenes, effectively compensating for data scarcity in 4D-PSG. Extensive experiments on the benchmark data demonstrate that we strikingly outperform baseline models by a large margin, highlighting the effectiveness of our method.

In this paper, we propose LSRNA, a novel framework for higher-resolution (exceeding 1K) image generation using diffusion models by leveraging super-resolution directly in the latent space. Existing diffusion models struggle with scaling beyond their training resolutions, often leading to structural distortions or content repetition. Reference-based methods address the issues by upsampling a low-resolution reference to guide higher-resolution generation. However, they face significant challenges: upsampling in latent space often causes manifold deviation, which degrades output quality. On the other hand, upsampling in RGB space tends to produce overly smoothed outputs. To overcome these limitations, LSRNA combines Latent space Super-Resolution (LSR) for manifold alignment and Region-wise Noise Addition (RNA) to enhance high-frequency details. Our extensive experiments demonstrate that integrating LSRNA outperforms state-of-the-art reference-based methods across various resolutions and metrics, while showing the critical role of latent space upsampling in preserving detail and sharpness. The code will be available at https://github.com/3587jjh/LSRNA https://github.com/3587jjh/LSRNA.

Customized text-to-image generation renders user-specified concepts into novel contexts based on textual prompts. Scaling the number of concepts in customized generation meets a broader demand for user creation, whereas existing methods face challenges with generation quality and computational efficiency. In this paper, we propose LaTexBlend, a novel framework for effectively and efficiently scaling multi-concept customized generation. The core idea of LaTexBlend is to represent single concepts and blend multiple concepts within a Latent Textual space, which is positioned after the text encoder and a linear projection. LaTexBlend customizes each concept individually, storing them in a concept bank with a compact representation of latent textual features that captures sufficient concept information to ensure high fidelity. At inference, concepts from the bank can be freely and seamlessly combined in the latent textual space, offering two key merits for multi-concept generation: 1) excellent scalability, and 2) significant reduction of denoising deviation, preserving coherent layouts. Extensive experiments demonstrate that LaTexBlend can flexibly integrate multiple customized concepts with harmonious structures and high subject fidelity, substantially outperforming baselines in both generation quality and computational efficiency. Our code will be publicly available.

Generating natural listener responses in conversational scenarios is crucial for creating engaging digital humans and avatars. Recent work has shown that large language models (LLMs) can be effectively leveraged for this task, demonstrating remarkable capabilities in generating contextually appropriate listener behaviors. However, current LLM-based methods face two critical limitations: they rely solely on speech content, overlooking other crucial communication signals, and they entangle listener identity with response generation, compromising output fidelity and generalization. In this work, we present a novel framework that addresses these limitations while maintaining the advantages of LLMs. Our approach introduces a Multimodal-LM architecture that jointly processes speech content, acoustics, and speaker emotion, capturing the full spectrum of communication cues. Additionally, we propose an identity disentanglement strategy using instance normalization and adaptive instance normalization in a VQ-VAE framework, enabling high-fidelity listening head synthesis with flexible identity control. Extensive experiments demonstrate that our method significantly outperforms existing approaches in terms of response naturalness and fidelity, while enabling effective identity control without retraining.

Most real-world image editing tasks require multiple sequential edits to achieve desired results. Current editing approaches, primarily designed for single-object modifications, struggle with sequential editing: especially with maintaining previous edits along with adapting new objects naturally into the existing content. These limitations significantly hinder complex editing scenarios where multiple objects need to be modified while preserving their contextual relationships. We address this fundamental challenge through two key proposals: enabling rough mask inputs that preserve existing content while naturally integrating new elements and supporting consistent editing across multiple modifications. Our framework achieves this through layer-wise memory, which stores latent representations and prompt embeddings from previous edits. We propose Background Consistency Guidance that leverages memorized latents to maintain scene coherence and Multi-Query Disentanglement in cross-attention that ensures natural adaptation to existing content. To evaluate our method, we present a new benchmark dataset incorporating semantic alignment metrics and interactive editing scenarios. Through comprehensive experiments, we demonstrate superior performance in iterative image editing tasks with minimal user effort, requiring only rough masks while maintaining high-quality results throughout multiple editing steps.

Employing LLMs for visual generation has recently become a research focus. However, the existing methods primarily transfer the LLM architecture to visual generation but rarely investigate the fundamental differences between language and vision. This oversight may lead to suboptimal utilization of visual generation capabilities within the LLM framework. In this paper, we explore the characteristics of visual embedding space under the LLM framework and discover that the correlation between visual embeddings can help achieve more stable and robust generation results. We present **IAR**, an **I**mproved **A**uto**R**egressive Visual Generation Method that enhances the training efficiency and generation quality of LLM-based visual generation models. Firstly, we propose a Codebook Rearrangement strategy that uses balanced k-means clustering algorithm to rearrange the visual codebook into clusters, ensuring high similarity among visual features within each cluster. Leveraging the rearranged codebook, we propose a Cluster-oriented Cross-entropy Loss that guides the model to correctly predict the cluster where the token is located. This approach ensures that even if the model predicts the wrong token index, there is a high probability the predicted token is located in the correct cluster, which significantly enhances the generation quality and robustness. Extensive experiments demonstrate that our method consistently enhances the model training efficiency and performance from 100M to 1.4B, reducing the training time by half while achieving the same FID. Additionally, our approach can be applied to various LLM-based visual generation models and adheres to the scaling law, providing a promising direction for future research in LLM-based visual generation.

In-domain generation aims to perform a variety of tasks within a specific domain, such as unconditional generation, text-to-image, image editing, 3D generation, and more. Early research typically required training specialized generators for each unique task and domain, often relying on fully-labeled data. Motivated by the powerful generative capabilities and broad applications of diffusion models, we are driven to explore leveraging label-free data to empower these models for in-domain generation.Fine-tuning a pre-trained generative model on domain data is an intuitive but challenging way and often requires complex manual hyper-parameter adjustments since the limited diversity of the training data can easily disrupt the model's original generative capabilities.To address this challenge, we propose a guidance-decoupled prior preservation mechanism to achieve high generative quality and controllability by image-only data, inspired by preserving the pre-trained model from a denoising guidance perspective.We decouple domain-related guidance from the conditional guidance used in classifier-free guidance mechanisms to preserve open-world control guidance and unconditional guidance from the pre-trained model. We further propose an efficient domain knowledge learning technique to train an additional text-free UNet copy to predict domain guidance.Besides, we theoretically illustrate a multi-guidance in-domain generation pipeline for a variety of generative tasks, leveraging multiple guidances from distinct diffusion models and conditions. Extensive experiments demonstrate the superiority of our method in domain-specific synthesis and its compatibility with various diffusion-based control methods and applications.

Generative methods have reached a level of quality that is almost indistinguishable from real data. However, while individual samples may appear unique, generative models often exhibit limitations in covering the full data distribution. Unlike quality issues, diversity problems within generative models are not easily detected by simply observing single images or generated datasets, which means we need a specific measure to assess the diversity of these models. In this paper, we draw attention to the current lack of diversity in generative models and the inability of common metrics to measure this. We achieve this by framing diversity as an image retrieval problem, where we measure how many real images can be retrieved using synthetic data as queries. This yields the Image Retrieval Score (IRS), an interpretable, hyperparameter-free metric that quantifies the diversity of a generative model's output. IRS requires only a subset of synthetic samples and provides a statistical measure of confidence. Our experiments indicate that current feature extractors commonly used in generative model assessment are inadequate for evaluating diversity effectively. Consequently, we perform an extensive search for the best feature extractors to assess diversity. Evaluation reveals that current diffusion models converge to limited subsets of the real distribution, with no current state-of-the-art models superpassing 77% of the diversity of the training data. To address this limitation, we introduce Diversity-Aware Diffusion Models (DiADM), a novel approach that improves diversity of unconditional diffusion models without loss of image quality. We do this by disentangling diversity from image quality by using a diversity aware module that uses pseudo-unconditional features as input. We provide a Python package offering unified feature extraction and metric computation to further facilitate the evaluation of generative models https://github.com/MischaD/beyondfid.

Multispectral (MS) images contain richer spectral information than RGB images due to their increased number of channels and are widely used for various applications. However, achieving accurate estimation in MS images remains challenging, as previous studies have struggled with spectral diversity and the inherent entanglement between the illuminant and surface reflectance spectra. To tackle these challenges, in this paper, we propose a novel Illumination spectrum estimation technique for MS images via Surface reflectance modeling and Spatial-spectral feature generation (ISS). The proposed technique employs a learnable spectral unmixing (SU) block to enhance surface reflectance modeling, which was unattempted in the illumination spectrum estimation, and a feature mixing block to fuse spectral and spatial features of MS images with cross-attention. The features are refined iteratively and processed through a decoder to produce an illumination spectrum estimator. Experimental results demonstrate that the proposed technique achieves state-of-the-art performance in illumination spectrum estimation in various MS image datasets. The code is available at https://github.com/heyjinnii/ISS-MSI.git.

Recently, zero-shot methods like InstantID have revolutionized identity-preserving generation. Unlike multi-image finetuning approaches such as DreamBooth, these zero-shot methods leverage powerful facial encoders to extract identity information from a single portrait photo, enabling efficient identity-preserving generation through a single inference pass. However, this convenience introduces new threats to the facial identity protection. This paper aims to safeguard portrait photos from unauthorized encoder-based customization. We introduce IDProtector, an adversarial noise encoder that applies imperceptible adversarial noise to portrait photos in a single forward pass. Our approach offers universal protection for portraits against multiple state-of-the-art encoder-based methods, including InstantID, IP-Adapter, and PhotoMaker, while ensuring robustness to common image transformations such as JPEG compression, resizing, and affine transformations. Experiments across diverse portrait datasets and generative models reveal that IDProtector generalizes effectively to unseen data and even closed-source proprietary models.

Personalized portrait synthesis, essential in domains like social entertainment, has recently made significant progress. Person-wise fine-tuning based methods, such as LoRA and DreamBooth, can produce photorealistic outputs but need training on individual samples, consuming time and resources and posing an unstable risk. Adapter based techniques such as IP-Adapter freeze the foundational model parameters and employ a plug-in architecture to enable zero-shot inference, but they often exhibit a lack of naturalness and authenticity, which are not to be overlooked in portrait synthesis tasks. In this paper, we introduce a parameter-efficient adaptive generation method, namely HyperLoRA, that uses an adaptive plug-in network to generate LoRA weights, merging the superior performance of LoRA with the zero-shot capability of adapter scheme. Through our carefully designed network structure and training strategy, we achieve zero-shot personalized portrait generation (supporting both single and multiple image inputs) with high photorealism, fidelity, and editability.

Multimodal LLMs have advanced vision-language tasks but still struggle with understanding video scenes. To bridge this gap, Video Scene Graph Generation (VidSGG) has emerged to capture multi-object relationships across video frames. However, prior methods rely on pairwise connections, limiting their ability to handle complex multi-object interactions and reasoning. To this end, we propose Multimodal LLMs on a Scene HyperGraph (HyperGLM), promoting reasoning about multi-way interactions and higher-order relationships. Our approach uniquely integrates entity scene graphs, which capture spatial relationships between objects, with a procedural graph that models their causal transitions, forming a unified HyperGraph. Significantly, HyperGLM enables reasoning by injecting this unified HyperGraph into LLMs. Additionally, we introduce a new Video Scene Graph Reasoning (VSGR) dataset featuring 1.9M frames from third-person, egocentric, and drone views and supports five tasks: Scene Graph Generation, Scene Graph Anticipation, Video Question Answering, Video Captioning, and Relation Reasoning. Empirically, HyperGLM consistently outperforms state-of-the-art methods across five tasks, effectively modeling and reasoning complex relationships in diverse video scenes.

Human-motion video generation has been a challenging task, primarily due to the difficulty inherent in learning human body movements. While some approaches have attempted to drive human-centric video generation explicitly through pose control, these methods typically rely on poses derived from existing videos, thereby lacking flexibility. To address this, we propose HumanDreamer, a decoupled human video generation framework that first generates diverse poses from text prompts and then leverages these poses to generate human-motion videos. Specifically, we propose MotionVid, the largest dataset for human-motion pose generation. Based on the dataset, we present MotionDiT, which is trained to generate structured human-motion poses from text prompts. Besides, a novel LAMA loss is introduced, which together contribute to a significant improvement in FID by 62.4%, along with respective enhancements in R-precision for top1, top2, and top3 by 41.8%, 26.3%, and 18.3%, thereby advancing both the Text-to-Pose control accuracy and FID metrics. Our experiments across various Pose-to-Video baselines demonstrate that the poses generated by our method can produce diverse and high-quality human-motion videos. Furthermore, our model can facilitate other downstream tasks, such as pose sequence prediction and 2D-3D motion lifting.

3D content creation has achieved significant progress in terms of both quality and speed. Although current Gaussian Splatting-based methods can produce 3D objects within seconds, they are still limited by complex preprocessing or low controllability. In this paper, we introduce a novel framework designed to efficiently and controllably generate high-resolution 3D models from text promptsor image. Our key insights are three-fold: 1) Hierarchical Gaussian Mixture Model Splatting: We propose an hybrid hierarchical representation to extract fixed number of fine-grained Gaussians with multiscale details from textured object, also establish part-level representation of Gaussians primitives. 2) Mamba with adaptive tree topology: We present a diffusion mamba with tree-topology to adaptively generate Gaussians with disordered spatial structures, without the need for complex preprocessing and maintain linear complexity generation. 3) Controllable Generation: Building on the HGMM tree, we introduce a cascaded diffusion framework combining controllable implicit latent generation, which progressively generates condition-driven latents, and explicit splatting generation, which transforms latents into high-quality Gaussian primitives. Extensive experiments demonstrate the high fidelity and efficiency of our approach.

Neural Radiance Fields (NeRF) have transformed novel view synthesis by modeling scene-specific volumetric representations directly from images. While generalizable NeRF models can generate novel views across unknown scenes by learning latent ray representations, their performance heavily depends on a large number of multi-view observations. However, with limited input views, these methods experience significant degradation in rendering quality. To address this limitation, we propose GoLF-NRT: a Global and Local feature Fusion-based Neural Rendering Transformer. GoLF-NRT enhances generalizable neural rendering from few input views by leveraging a 3D transformer with efficient sparse attention to capture global scene context. In parallel, it integrates local geometric features extracted along the epipolar line, enabling high-quality scene reconstruction from as few as 1 to 3 input views. Furthermore, we introduce an adaptive sampling strategy based on attention weights and kernel regression, improving the accuracy of transformer-based neural rendering. Extensive experiments on public datasets show that GoLF-NRT achieves state-of-the-art performance across varying numbers of input views, highlighting the effectiveness and superiority of our approach. Code is available at https://github.com/KLMAV-CUC/GoLF-NRT.

Large Vision-Language Models (VLMs), such as GPT-4, have achieved remarkable success across various fields. However, there are few studies on 3D indoor scene generation with VLMs. This paper considers this task as a planning problem subject to spatial and layout common sense constraints. To solve the problem with a VLM, we propose a new global-local tree search algorithm. Globally, the method places each object sequentially and explores multiple placements during each placement process, where the problem space is represented as a tree. To reduce the depth of the tree, we decompose the scene structure hierarchically, i.e. room level, region level, floor object level, and supported object level. The algorithm independently generates the floor objects in different regions and supported objects placed on different floor objects. Locally, we also decompose the sub-task, the placement of each object, into multiple steps. The algorithm searches the tree of problem space. To leverage the VLM model to produce positions of objects, we discretize the top-down view space as a dense grid and fill each cell with diverse emojis to make to cells distinct. We prompt the VLM with the emoji grid and the VLM produces a reasonable location for the object by describing the position with the name of emojis. The quantitative and qualitative experimental results illustrate our approach generates more plausible 3D scenes than state-of-the-art approaches. Our source code is available at https://github.com/dw-dengwei/TreeSearchGen .

Generative inbetweening aims to generate intermediate frame sequences by utilizing two key frames as input. Although remarkable progress has been made in video generation models, generative inbetweening still faces challenges in maintaining temporal stability due to the ambiguous interpolation path between two key frames. This issue becomes particularly severe when there is a large motion gap between input frames. In this paper, we propose a straightforward yet highly effective Frame-wise Conditions-driven Video Generation (FCVG) method that significantly enhances the temporal stability of interpolated video frames. Specifically, our FCVG provides an explicit condition for each frame, making it much easier to identify the interpolation path between two input frames and thus ensuring temporally stable production of visually plausible video frames. To achieve this, we suggest extracting matched lines from two input frames that can then be easily interpolated frame by frame, serving as frame-wise conditions seamlessly integrated into existing video generation models. In extensive evaluations covering diverse scenarios such as natural landscapes, complex human poses, camera movements and animations, existing methods often exhibit incoherent transitions across frames. In contrast, our FCVG demonstrates the capability to generate temporally stable videos using both linear and non-linear interpolation curves. Our project page and code are available at https://fcvg-inbetween.github.io/.

We show that the GPS tags contained in photo metadata provide a useful control signal for image generation. We train GPS-to-image models and use them for tasks that require a fine-grained understanding of how images vary within a city. In particular, we train a diffusion model to generate images conditioned on both GPS and text. The learned model generates images that capture the distinctive appearance of different neighborhoods, parks, and landmarks. We also extract 3D models from 2D GPS-to-image models through score distillation sampling, using GPS conditioning to constrain the appearance of the reconstruction from each viewpoint. Our evaluations suggest that our GPS-conditioned models successfully learn to generate images that vary based on location, and that GPS conditioning improves estimated 3D structure.