arXiv subjects

Guanglu Song

Publications and source records attributed to Guanglu Song.

At least 19 recordsLinked to original sources

arXiv2026

Omni-LiveAvatar: Minute-Level Real-Time Streaming Joint Audio-Video Avatar Generation

Joint audio-video generative models serve as foundation for immersive and interactive digital-human generation. Nevertheless, most existing models rely on bidirectional attention and multi-step denoising and can generate only short clips, making them unsuitable for real-time interaction over extended durations. We present Omni-LiveAvatar, the first framework for minute-level, real-time streaming joint audio-video avatar generation. Specifically, we propose (1) a progressive autoregressive distillation pipeline that transfers a large bidirectional joint audio-video diffusion model into a few-step autoregressive generator without auxiliary stabilization mechanisms; (2) a synchronized audio-video long-short-term memory that preserves global consistency under a bounded memory budget; and (3) a hierarchical rolling prompt planning strategy that enables coherent semantic evolution and seamless prompt transitions. Extensive experiments show that Omni-LiveAvatar generates high-quality, synchronized minute-level avatars in real time. In terms of speed, it achieves a 33$\times$ generation speedup over its teacher, LTX-2, on a single NVIDIA H200 GPU; in terms of generation quality, it outperforms accelerated baselines across visual quality, audio quality, cross-modal synchronization, and human fidelity. Our code is available at https://github.com/Aoko955/Omni-LiveAvatar.

cs.MM↗

arXiv2026

Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation

Distilling video generation models to extremely low inference budgets (e.g., 2--4 NFEs) is crucial for real-time deployment, yet remains challenging. Trajectory-style consistency distillation often becomes conservative under complex video dynamics, yielding an over-smoothed appearance and weak motion. Distribution matching distillation (DMD) can recover sharp, mode-seeking samples, but its local training signals do not explicitly regularize how denoising updates compose across timesteps, making composed rollouts prone to drift. To overcome this challenge, we propose Self-Consistent Distribution Matching Distillation (SC-DMD), which explicitly regularizes the endpoint-consistent composition of consecutive denoising updates. For real-time autoregressive video generation, we further treat the KV cache as a quality parameterized condition and propose Cache-Distribution-Aware training. This training scheme applies SC-DMD over multi-step rollouts and introduces a cache-conditioned feature alignment objective that steers low-quality outputs toward high-quality references. Across extensive experiments on both non-autoregressive backbones (e.g., Wan~2.1) and autoregressive real-time paradigms (e.g., Self Forcing), our method, dubbed \textbf{Salt}, consistently improves low-NFE video generation quality while remaining compatible with diverse KV-cache memory mechanisms. Project page: https://xingtongge.github.io/Salt

cs.CV↗

arXiv2026

Improving Joint Audio-Video Generation with Cross-Modal Context Learning

The dual-stream transformer architecture-based joint audio-video generation method has become the dominant paradigm in current research. By incorporating pre-trained video diffusion models and audio diffusion models, along with a cross-modal interaction attention module, high-quality, temporally synchronized audio-video content can be generated with minimal training data. In this paper, we first revisit the dual-stream transformer paradigm and further analyze its limitations, including model manifold variations caused by the gating mechanism controlling cross-modal interactions, biases in multi-modal background regions introduced by cross-modal attention, and the inconsistencies in multi-modal classifier-free guidance (CFG) during training and inference, as well as conflicts between multiple conditions. To alleviate these issues, we propose Cross-Modal Context Learning (CCL), equipped with several carefully designed modules. Temporally Aligned RoPE and Partitioning (TARP) effectively enhances the temporal alignment between audio latent and video latent representations. The Learnable Context Tokens (LCT) and Dynamic Context Routing (DCR) in the Cross-Modal Context Attention (CCA) module provide stable unconditional anchors for cross-modal information, while dynamically routing based on different training tasks, further enhancing the model's convergence speed and generation quality. During inference, Unconditional Context Guidance (UCG) leverages the unconditional support provided by LCT to facilitate different forms of CFG, improving train-inference consistency and further alleviating conflicts. Through comprehensive evaluations, CCL achieves state-of-the-art performance compared with recent academic methods while requiring substantially fewer resources.

cs.CV↗

arXiv2026

AR-CoPO: Align Autoregressive Video Generation with Contrastive Policy Optimization

Streaming autoregressive (AR) video generators combined with few-step distillation achieve low-latency, high-quality synthesis, yet remain difficult to align via reinforcement learning from human feedback (RLHF). Existing SDE-based GRPO methods face challenges in this setting: few-step ODEs and consistency model samplers deviate from standard flow-matching ODEs, and their short, low-stochasticity trajectories are highly sensitive to initialization noise, rendering intermediate SDE exploration ineffective. We propose AR-CoPO (AutoRegressive Contrastive Policy Optimization), a framework that adapts the Neighbor GRPO contrastive perspective to streaming AR generation. AR-CoPO introduces chunk-level alignment via a forking mechanism that constructs neighborhood candidates at a randomly selected chunk, assigns sequence-level rewards, and performs localized GRPO updates. We further propose a semi-on-policy training strategy that complements on-policy exploration with exploitation over a replay buffer of reference rollouts, improving generation quality across domains. Experiments on Self-Forcing demonstrate that AR-CoPO improves both out-of-domain generalization and in-domain human preference alignment over the baseline, providing evidence of genuine alignment rather than reward hacking.

cs.CV↗

arXiv2025

Neighbor GRPO: Contrastive ODE Policy Optimization Aligns Flow Models

Group Relative Policy Optimization (GRPO) has shown promise in aligning image and video generative models with human preferences. However, applying it to modern flow matching models is challenging because of its deterministic sampling paradigm. Current methods address this issue by converting Ordinary Differential Equations (ODEs) to Stochastic Differential Equations (SDEs), which introduce stochasticity. However, this SDE-based GRPO suffers from issues of inefficient credit assignment and incompatibility with high-order solvers for fewer-step sampling. In this paper, we first reinterpret existing SDE-based GRPO methods from a distance optimization perspective, revealing their underlying mechanism as a form of contrastive learning. Based on this insight, we propose Neighbor GRPO, a novel alignment algorithm that completely bypasses the need for SDEs. Neighbor GRPO generates a diverse set of candidate trajectories by perturbing the initial noise conditions of the ODE and optimizes the model using a softmax distance-based surrogate leaping policy. We establish a theoretical connection between this distance-based objective and policy gradient optimization, rigorously integrating our approach into the GRPO framework. Our method fully preserves the advantages of deterministic ODE sampling, including efficiency and compatibility with high-order solvers. We further introduce symmetric anchor sampling for computational efficiency and group-wise quasi-norm reweighting to address reward flattening. Extensive experiments demonstrate that Neighbor GRPO significantly outperforms SDE-based counterparts in terms of training cost, convergence speed, and generation quality.

cs.CV↗

arXiv2025

Towards Seamless Borders: A Method for Mitigating Inconsistencies in Image Inpainting and Outpainting

Image inpainting is the task of reconstructing missing or damaged parts of an image in a way that seamlessly blends with the surrounding content. With the advent of advanced generative models, especially diffusion models and generative adversarial networks, inpainting has achieved remarkable improvements in visual quality and coherence. However, achieving seamless continuity remains a significant challenge. In this work, we propose two novel methods to address discrepancy issues in diffusion-based inpainting models. First, we introduce a modified Variational Autoencoder that corrects color imbalances, ensuring that the final inpainted results are free of color mismatches. Second, we propose a two-step training strategy that improves the blending of generated and existing image content during the diffusion process. Through extensive experiments, we demonstrate that our methods effectively reduce discontinuity and produce high-quality inpainting results that are coherent and visually appealing.

cs.CV↗

arXiv2025

ADT: Tuning Diffusion Models with Adversarial Supervision

Diffusion models have achieved outstanding image generation by reversing a forward noising process to approximate true data distributions. During training, these models predict diffusion scores from noised versions of true samples in a single forward pass, while inference requires iterative denoising starting from white noise. This training-inference divergences hinder the alignment between inference and training data distributions, due to potential prediction biases and cumulative error accumulation. To address this problem, we propose an intuitive but effective fine-tuning framework, called Adversarial Diffusion Tuning (ADT), by stimulating the inference process during optimization and aligning the final outputs with training data by adversarial supervision. Specifically, to achieve robust adversarial training, ADT features a siamese-network discriminator with a fixed pre-trained backbone and lightweight trainable parameters, incorporates an image-to-image sampling strategy to smooth discriminative difficulties, and preserves the original diffusion loss to prevent discriminator hacking. In addition, we carefully constrain the backward-flowing path for back-propagating gradients along the inference path without incurring memory overload or gradient explosion. Finally, extensive experiments on Stable Diffusion models (v1.5, XL, and v3), demonstrate that ADT significantly improves both distribution alignment and image quality.

cs.CV↗

arXiv2025

High-Fidelity Diffusion Face Swapping with ID-Constrained Facial Conditioning

Face swapping aims to seamlessly transfer a source facial identity onto a target while preserving target attributes such as pose and expression. Diffusion models, known for their superior generative capabilities, have recently shown promise in advancing face-swapping quality. This paper addresses two key challenges in diffusion-based face swapping: the prioritized preservation of identity over target attributes and the inherent conflict between identity and attribute conditioning. To tackle these issues, we introduce an identity-constrained attribute-tuning framework for face swapping that first ensures identity preservation and then fine-tunes for attribute alignment, achieved through a decoupled condition injection. We further enhance fidelity by incorporating identity and adversarial losses in a post-training refinement stage. Our proposed identity-constrained diffusion-based face-swapping model outperforms existing methods in both qualitative and quantitative evaluations, demonstrating superior identity similarity and attribute consistency, achieving a new state-of-the-art performance in high-fidelity face swapping.

cs.CV↗

arXiv2024

VividFace: A Diffusion-Based Hybrid Framework for High-Fidelity Video Face Swapping

Video face swapping is becoming increasingly popular across various applications, yet existing methods primarily focus on static images and struggle with video face swapping because of temporal consistency and complex scenarios. In this paper, we present the first diffusion-based framework specifically designed for video face swapping. Our approach introduces a novel image-video hybrid training framework that leverages both abundant static image data and temporal video sequences, addressing the inherent limitations of video-only training. The framework incorporates a specially designed diffusion model coupled with a VidFaceVAE that effectively processes both types of data to better maintain temporal coherence of the generated videos. To further disentangle identity and pose features, we construct the Attribute-Identity Disentanglement Triplet (AIDT) Dataset, where each triplet has three face images, with two images sharing the same pose and two sharing the same identity. Enhanced with a comprehensive occlusion augmentation, this dataset also improves robustness against occlusions. Additionally, we integrate 3D reconstruction techniques as input conditioning to our network for handling large pose variations. Extensive experiments demonstrate that our framework achieves superior performance in identity preservation, temporal consistency, and visual quality compared to existing methods, while requiring fewer inference steps. Our approach effectively mitigates key challenges in video face swapping, including temporal flickering, identity preservation, and robustness to occlusions and pose variations.

cs.CV↗

arXiv2024

EasyRef: Omni-Generalized Group Image Reference for Diffusion Models via Multimodal LLM

Significant achievements in personalization of diffusion models have been witnessed. Conventional tuning-free methods mostly encode multiple reference images by averaging their image embeddings as the injection condition, but such an image-independent operation cannot perform interaction among images to capture consistent visual elements within multiple references. Although the tuning-based Low-Rank Adaptation (LoRA) can effectively extract consistent elements within multiple images through the training process, it necessitates specific finetuning for each distinct image group. This paper introduces EasyRef, a novel plug-and-play adaptation method that enables diffusion models to be conditioned on multiple reference images and the text prompt. To effectively exploit consistent visual elements within multiple images, we leverage the multi-image comprehension and instruction-following capabilities of the multimodal large language model (MLLM), prompting it to capture consistent visual elements based on the instruction. Besides, injecting the MLLM's representations into the diffusion process through adapters can easily generalize to unseen domains, mining the consistent visual elements within unseen data. To mitigate computational costs and enhance fine-grained detail preservation, we introduce an efficient reference aggregation strategy and a progressive training scheme. Finally, we introduce MRBench, a new multi-reference image generation benchmark. Experimental results demonstrate EasyRef surpasses both tuning-free methods like IP-Adapter and tuning-based methods like LoRA, achieving superior aesthetic quality and robust zero-shot generalization across diverse domains.

cs.CV↗

arXiv2024

See Further When Clear: Curriculum Consistency Model

Significant advances have been made in the sampling efficiency of diffusion models and flow matching models, driven by Consistency Distillation (CD), which trains a student model to mimic the output of a teacher model at a later timestep. However, we found that the learning complexity of the student model varies significantly across different timesteps, leading to suboptimal performance in CD.To address this issue, we propose the Curriculum Consistency Model (CCM), which stabilizes and balances the learning complexity across timesteps. Specifically, we regard the distillation process at each timestep as a curriculum and introduce a metric based on Peak Signal-to-Noise Ratio (PSNR) to quantify the learning complexity of this curriculum, then ensure that the curriculum maintains consistent learning complexity across different timesteps by having the teacher model iterate more steps when the noise intensity is low. Our method achieves competitive single-step sampling Fr\'echet Inception Distance (FID) scores of 1.64 on CIFAR-10 and 2.18 on ImageNet 64x64.Moreover, we have extended our method to large-scale text-to-image models and confirmed that it generalizes well to both diffusion models (Stable Diffusion XL) and flow matching models (Stable Diffusion 3). The generated samples demonstrate improved image-text alignment and semantic structure, since CCM enlarges the distillation step at large timesteps and reduces the accumulated error.

cs.CV↗

arXiv2024

Robo-MUTUAL: Robotic Multimodal Task Specification via Unimodal Learning

Multimodal task specification is essential for enhanced robotic performance, where \textit{Cross-modality Alignment} enables the robot to holistically understand complex task instructions. Directly annotating multimodal instructions for model training proves impractical, due to the sparsity of paired multimodal data. In this study, we demonstrate that by leveraging unimodal instructions abundant in real data, we can effectively teach robots to learn multimodal task specifications. First, we endow the robot with strong \textit{Cross-modality Alignment} capabilities, by pretraining a robotic multimodal encoder using extensive out-of-domain data. Then, we employ two Collapse and Corrupt operations to further bridge the remaining modality gap in the learned multimodal representation. This approach projects different modalities of identical task goal as interchangeable representations, thus enabling accurate robotic operations within a well-aligned multimodal latent space. Evaluation across more than 130 tasks and 4000 evaluations on both simulated LIBERO benchmark and real robot platforms showcases the superior capabilities of our proposed framework, demonstrating significant advantage in overcoming data constraints in robotic learning. Website: zh1hao.wang/Robo_MUTUAL

cs.RO↗

arXiv2024

MMSearch: Benchmarking the Potential of Large Models as Multi-modal Search Engines

The advent of Large Language Models (LLMs) has paved the way for AI search engines, e.g., SearchGPT, showcasing a new paradigm in human-internet interaction. However, most current AI search engines are limited to text-only settings, neglecting the multimodal user queries and the text-image interleaved nature of website information. Recently, Large Multimodal Models (LMMs) have made impressive strides. Yet, whether they can function as AI search engines remains under-explored, leaving the potential of LMMs in multimodal search an open question. To this end, we first design a delicate pipeline, MMSearch-Engine, to empower any LMMs with multimodal search capabilities. On top of this, we introduce MMSearch, a comprehensive evaluation benchmark to assess the multimodal search performance of LMMs. The curated dataset contains 300 manually collected instances spanning 14 subfields, which involves no overlap with the current LMMs' training data, ensuring the correct answer can only be obtained within searching. By using MMSearch-Engine, the LMMs are evaluated by performing three individual tasks (requery, rerank, and summarization), and one challenging end-to-end task with a complete searching process. We conduct extensive experiments on closed-source and open-source LMMs. Among all tested models, GPT-4o with MMSearch-Engine achieves the best results, which surpasses the commercial product, Perplexity Pro, in the end-to-end task, demonstrating the effectiveness of our proposed pipeline. We further present error analysis to unveil current LMMs still struggle to fully grasp the multimodal search tasks, and conduct ablation study to indicate the potential of scaling test-time computation for AI search engine. We hope MMSearch may provide unique insights to guide the future development of multimodal AI search engine. Project Page: https://mmsearch.github.io

cs.CV↗

arXiv2024

Exploring the Role of Large Language Models in Prompt Encoding for Diffusion Models

Large language models (LLMs) based on decoder-only transformers have demonstrated superior text understanding capabilities compared to CLIP and T5-series models. However, the paradigm for utilizing current advanced LLMs in text-to-image diffusion models remains to be explored. We observed an unusual phenomenon: directly using a large language model as the prompt encoder significantly degrades the prompt-following ability in image generation. We identified two main obstacles behind this issue. One is the misalignment between the next token prediction training in LLM and the requirement for discriminative prompt features in diffusion models. The other is the intrinsic positional bias introduced by the decoder-only architecture. To deal with this issue, we propose a novel framework to fully harness the capabilities of LLMs. Through the carefully designed usage guidance, we effectively enhance the text representation capability for prompt encoding and eliminate its inherent positional bias. This allows us to integrate state-of-the-art LLMs into the text-to-image generation model flexibly. Furthermore, we also provide an effective manner to fuse multiple LLMs into our framework. Considering the excellent performance and scaling capabilities demonstrated by the transformer architecture, we further design an LLM-Infused Diffusion Transformer (LI-DiT) based on the framework. We conduct extensive experiments to validate LI-DiT across model size and data size. Benefiting from the inherent ability of the LLMs and our innovative designs, the prompt understanding performance of LI-DiT easily surpasses state-of-the-art open-source models as well as mainstream closed-source commercial models including Stable Diffusion 3, DALL-E 3, and Midjourney V6. The LLM-Infused Diffuser framework is also one of the core technologies powering SenseMirage, a highly advanced text-to-image model.

cs.CV↗

arXiv2024

Phased Consistency Models

Consistency Models (CMs) have made significant progress in accelerating the generation of diffusion models. However, their application to high-resolution, text-conditioned image generation in the latent space remains unsatisfactory. In this paper, we identify three key flaws in the current design of Latent Consistency Models (LCMs). We investigate the reasons behind these limitations and propose Phased Consistency Models (PCMs), which generalize the design space and address the identified limitations. Our evaluations demonstrate that PCMs outperform LCMs across 1--16 step generation settings. While PCMs are specifically designed for multi-step refinement, they achieve comparable 1-step generation results to previously state-of-the-art specifically designed 1-step methods. Furthermore, we show the methodology of PCMs is versatile and applicable to video generation, enabling us to train the state-of-the-art few-step text-to-video generator. Our code is available at https://github.com/G-U-N/Phased-Consistency-Model.

cs.LG↗

arXiv2024

Deep Reward Supervisions for Tuning Text-to-Image Diffusion Models

Optimizing a text-to-image diffusion model with a given reward function is an important but underexplored research area. In this study, we propose Deep Reward Tuning (DRTune), an algorithm that directly supervises the final output image of a text-to-image diffusion model and back-propagates through the iterative sampling process to the input noise. We find that training earlier steps in the sampling process is crucial for low-level rewards, and deep supervision can be achieved efficiently and effectively by stopping the gradient of the denoising network input. DRTune is extensively evaluated on various reward models. It consistently outperforms other algorithms, particularly for low-level control signals, where all shallow supervision methods fail. Additionally, we fine-tune Stable Diffusion XL 1.0 (SDXL 1.0) model via DRTune to optimize Human Preference Score v2.1, resulting in the Favorable Diffusion XL 1.0 (FDXL 1.0) model. FDXL 1.0 significantly enhances image quality compared to SDXL 1.0 and reaches comparable quality compared with Midjourney v5.2.

cs.CV↗

arXiv2024

MoVA: Adapting Mixture of Vision Experts to Multimodal Context

As the key component in multimodal large language models (MLLMs), the ability of the visual encoder greatly affects MLLM's understanding on diverse image content. Although some large-scale pretrained vision encoders such as vision encoders in CLIP and DINOv2 have brought promising performance, we found that there is still no single vision encoder that can dominate various image content understanding, e.g., the CLIP vision encoder leads to outstanding results on general image understanding but poor performance on document or chart content. To alleviate the bias of CLIP vision encoder, we first delve into the inherent behavior of different pre-trained vision encoders and then propose the MoVA, a powerful and novel MLLM, adaptively routing and fusing task-specific vision experts with a coarse-to-fine mechanism. In the coarse-grained stage, we design a context-aware expert routing strategy to dynamically select the most suitable vision experts according to the user instruction, input image, and expertise of vision experts. This benefits from the powerful model function understanding ability of the large language model (LLM). In the fine-grained stage, we elaborately conduct the mixture-of-vision-expert adapter (MoV-Adapter) to extract and fuse task-specific knowledge from various experts. This coarse-to-fine paradigm effectively leverages representations from experts based on multimodal context and model expertise, further enhancing the generalization ability. We conduct extensive experiments to evaluate the effectiveness of the proposed approach. Without any bells and whistles, MoVA can achieve significant performance gains over current state-of-the-art methods in a wide range of challenging multimodal benchmarks.

cs.CV↗

arXiv2024

Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidance

Classifier-Free Guidance (CFG) has been widely used in text-to-image diffusion models, where the CFG scale is introduced to control the strength of text guidance on the whole image space. However, we argue that a global CFG scale results in spatial inconsistency on varying semantic strengths and suboptimal image quality. To address this problem, we present a novel approach, Semantic-aware Classifier-Free Guidance (S-CFG), to customize the guidance degrees for different semantic units in text-to-image diffusion models. Specifically, we first design a training-free semantic segmentation method to partition the latent image into relatively independent semantic regions at each denoising step. In particular, the cross-attention map in the denoising U-net backbone is renormalized for assigning each patch to the corresponding token, while the self-attention map is used to complete the semantic regions. Then, to balance the amplification of diverse semantic units, we adaptively adjust the CFG scales across different semantic regions to rescale the text guidance degrees into a uniform level. Finally, extensive experiments demonstrate the superiority of S-CFG over the original CFG strategy on various text-to-image diffusion models, without requiring any extra training cost. our codes are available at https://github.com/SmilesDZgk/S-CFG.

cs.CV↗