Pan Xie

Ying Guo, Qijun Gan, Yifu Zhang et al.

Video generation is rapidly evolving towards unified audio-video generation. In this paper, we present ALIVE, a generation model that adapts a pretrained Text-to-Video (T2V) model to Sora-style audio-video generation and animation. In particular, the model unlocks the Text-to-Video&Audio (T2VA) and Reference-to-Video&Audio (animation) capabilities compared to the T2V foundation models. To support the audio-visual synchronization and reference animation, we augment the popular MMDiT architecture with a joint audio-video branch which includes TA-CrossAttn for temporally-aligned cross-modal fusion and UniTemp-RoPE for precise audio-visual alignment. Meanwhile, a comprehensive data pipeline consisting of audio-video captioning, quality control, etc., is carefully designed to collect high-quality finetuning data. Additionally, we introduce a new benchmark to perform a comprehensive model test and comparison. After continue pretraining and finetuning on million-level high-quality data, ALIVE demonstrates outstanding performance, consistently outperforming open-source models and matching or surpassing state-of-the-art commercial solutions. With detailed recipes and benchmarks, we hope ALIVE helps the community develop audio-video generation models more efficiently. Official page: https://github.com/FoundationVision/Alive.

Pan Xie, Qipeng Zhang, Taiyi Peng et al.

The Sign Language Production (SLP) project aims to automatically translate spoken languages into sign sequences. Our approach focuses on the transformation of sign gloss sequences into their corresponding sign pose sequences (G2P). In this paper, we present a novel solution for this task by converting the continuous pose space generation problem into a discrete sequence generation problem. We introduce the Pose-VQVAE framework, which combines Variational Autoencoders (VAEs) with vector quantization to produce a discrete latent representation for continuous pose sequences. Additionally, we propose the G2P-DDM model, a discrete denoising diffusion architecture for length-varied discrete sequence data, to model the latent prior. To further enhance the quality of pose sequence generation in the discrete space, we present the CodeUnet model to leverage spatial-temporal information. Lastly, we develop a heuristic sequential clustering method to predict variable lengths of pose sequences for corresponding gloss sequences. Our results show that our model outperforms state-of-the-art G2P models on the public SLP evaluation benchmark. For more generated results, please visit our project page: \textcolor{blue}{\url{https://slpdiffusier.github.io/g2p-ddm}}

Yuxi Ren, Xin Xia, Yanzuo Lu et al.

Recently, a series of diffusion-aware distillation algorithms have emerged to alleviate the computational overhead associated with the multi-step inference process of Diffusion Models (DMs). Current distillation techniques often dichotomize into two distinct aspects: i) ODE Trajectory Preservation; and ii) ODE Trajectory Reformulation. However, these approaches suffer from severe performance degradation or domain shifts. To address these limitations, we propose Hyper-SD, a novel framework that synergistically amalgamates the advantages of ODE Trajectory Preservation and Reformulation, while maintaining near-lossless performance during step compression. Firstly, we introduce Trajectory Segmented Consistency Distillation to progressively perform consistent distillation within pre-defined time-step segments, which facilitates the preservation of the original ODE trajectory from a higher-order perspective. Secondly, we incorporate human feedback learning to boost the performance of the model in a low-step regime and mitigate the performance loss incurred by the distillation process. Thirdly, we integrate score distillation to further improve the low-step generation capability of the model and offer the first attempt to leverage a unified LoRA to support the inference process at all steps. Extensive experiments and user studies demonstrate that Hyper-SD achieves SOTA performance from 1 to 8 inference steps for both SDXL and SD1.5. For example, Hyper-SDXL surpasses SDXL-Lightning by +0.68 in CLIP Score and +0.51 in Aes Score in the 1-step inference.

Qijun Gan, Yi Ren, Chen Zhang et al.

Human motion video generation has advanced significantly, while existing methods still struggle with accurately rendering detailed body parts like hands and faces, especially in long sequences and intricate motions. Current approaches also rely on fixed resolution and struggle to maintain visual consistency. To address these limitations, we propose HumanDiT, a pose-guided Diffusion Transformer (DiT)-based framework trained on a large and wild dataset containing 14,000 hours of high-quality video to produce high-fidelity videos with fine-grained body rendering. Specifically, (i) HumanDiT, built on DiT, supports numerous video resolutions and variable sequence lengths, facilitating learning for long-sequence video generation; (ii) we introduce a prefix-latent reference strategy to maintain personalized characteristics across extended sequences. Furthermore, during inference, HumanDiT leverages Keypoint-DiT to generate subsequent pose sequences, facilitating video continuation from static images or existing videos. It also utilizes a Pose Adapter to enable pose transfer with given sequences. Extensive experiments demonstrate its superior performance in generating long-form, pose-accurate videos across diverse scenarios.

Jiaxiang Cheng, Pan Xie, Xin Xia et al.

Recent advancement in text-to-image models (e.g., Stable Diffusion) and corresponding personalized technologies (e.g., DreamBooth and LoRA) enables individuals to generate high-quality and imaginative images. However, they often suffer from limitations when generating images with resolutions outside of their trained domain. To overcome this limitation, we present the Resolution Adapter (ResAdapter), a domain-consistent adapter designed for diffusion models to generate images with unrestricted resolutions and aspect ratios. Unlike other multi-resolution generation methods that process images of static resolution with complex post-process operations, ResAdapter directly generates images with the dynamical resolution. Especially, after learning a deep understanding of pure resolution priors, ResAdapter trained on the general dataset, generates resolution-free images with personalized diffusion models while preserving their original style domain. Comprehensive experiments demonstrate that ResAdapter with only 0.5M can process images with flexible resolutions for arbitrary diffusion models. More extended experiments demonstrate that ResAdapter is compatible with other modules (e.g., ControlNet, IP-Adapter and LCM-LoRA) for image generation across a broad range of resolutions, and can be integrated into other multi-resolution model (e.g., ElasticDiffusion) for efficiently generating higher-resolution images. Project link is https://res-adapter.github.io

Yuxi Ren, Jie Wu, Yanzuo Lu et al.

Recent advancements in diffusion-based generative image editing have sparked a profound revolution, reshaping the landscape of image outpainting and inpainting tasks. Despite these strides, the field grapples with inherent challenges, including: i) inferior quality; ii) poor consistency; iii) insufficient instrcution adherence; iv) suboptimal generation efficiency. To address these obstacles, we present ByteEdit, an innovative feedback learning framework meticulously designed to Boost, Comply, and Accelerate Generative Image Editing tasks. ByteEdit seamlessly integrates image reward models dedicated to enhancing aesthetics and image-text alignment, while also introducing a dense, pixel-level reward model tailored to foster coherence in the output. Furthermore, we propose a pioneering adversarial and progressive feedback learning strategy to expedite the model's inference speed. Through extensive large-scale user evaluations, we demonstrate that ByteEdit surpasses leading generative image editing products, including Adobe, Canva, and MeiTu, in both generation quality and consistency. ByteEdit-Outpainting exhibits a remarkable enhancement of 388% and 135% in quality and consistency, respectively, when compared to the baseline model. Experiments also verfied that our acceleration models maintains excellent performance results in terms of quality and consistency.

Pan Xie, Taiyi Peng, Yao Du et al.

Sign Language Production (SLP) is the tough task of turning sign language into sign videos. The main goal of SLP is to create these videos using a sign gloss. In this research, we've developed a new method to make high-quality sign videos without using human poses as a middle step. Our model works in two main parts: first, it learns from a generator and the video's hidden features, and next, it uses another model to understand the order of these hidden features. To make this method even better for sign videos, we make several significant improvements. (i) In the first stage, we take an improved 3D VQ-GAN to learn downsampled latent representations. (ii) In the second stage, we introduce sequence-to-sequence attention to better leverage conditional information. (iii) The separated two-stage training discards the realistic visual semantic of the latent codes in the second stage. To endow the latent sequences semantic information, we extend the token-level autoregressive latent codes learning with perceptual loss and reconstruction loss for the prior model with visual perception. Compared with previous state-of-the-art approaches, our model performs consistently better on two word-level sign language datasets, i.e., WLASL and NMFs-CSL.

Jiacheng Zhang, Jie Wu, Yuxi Ren et al.

Latent diffusion models (LDM) have revolutionized text-to-image generation, leading to the proliferation of various advanced models and diverse downstream applications. However, despite these significant advancements, current diffusion models still suffer from several limitations, including inferior visual quality, inadequate aesthetic appeal, and inefficient inference, without a comprehensive solution in sight. To address these challenges, we present UniFL, a unified framework that leverages feedback learning to enhance diffusion models comprehensively. UniFL stands out as a universal, effective, and generalizable solution applicable to various diffusion models, such as SD1.5 and SDXL. Notably, UniFL consists of three key components: perceptual feedback learning, which enhances visual quality; decoupled feedback learning, which improves aesthetic appeal; and adversarial feedback learning, which accelerates inference. In-depth experiments and extensive user studies validate the superior performance of our method in enhancing generation quality and inference acceleration. For instance, UniFL surpasses ImageReward by 17% user preference in terms of generation quality and outperforms LCM and SDXL Turbo by 57% and 20% general preference with 4-step inference.

Hengnian Gu, Zhiyi Duan, Pan Xie et al.

The knowledge concept recommendation in Massive Open Online Courses (MOOCs) is a significant issue that has garnered widespread attention. Existing methods primarily rely on the explicit relations between users and knowledge concepts on the MOOC platforms for recommendation. However, there are numerous implicit relations (e.g., shared interests or same knowledge levels between users) generated within the users' learning activities on the MOOC platforms. Existing methods fail to consider these implicit relations, and these relations themselves are difficult to learn and represent, causing poor performance in knowledge concept recommendation and an inability to meet users' personalized needs. To address this issue, we propose a novel framework based on contrastive learning, which can represent and balance the explicit and implicit relations for knowledge concept recommendation in MOOCs (CL-KCRec). Specifically, we first construct a MOOCs heterogeneous information network (HIN) by modeling the data from the MOOC platforms. Then, we utilize a relation-updated graph convolutional network and stacked multi-channel graph neural network to represent the explicit and implicit relations in the HIN, respectively. Considering that the quantity of explicit relations is relatively fewer compared to implicit relations in MOOCs, we propose a contrastive learning with prototypical graph to enhance the representations of both relations to capture their fruitful inherent relational knowledge, which can guide the propagation of students' preferences within the HIN. Based on these enhanced representations, to ensure the balanced contribution of both towards the final recommendation, we propose a dual-head attention mechanism for balanced fusion. Experimental results demonstrate that CL-KCRec outperforms several state-of-the-art baselines on real-world datasets in terms of HR, NDCG and MRR.

Team Seedance, Heyi Chen, Siyan Chen et al.

Recent strides in video generation have paved the way for unified audio-visual generation. In this work, we present Seedance 1.5 pro, a foundational model engineered specifically for native, joint audio-video generation. Leveraging a dual-branch Diffusion Transformer architecture, the model integrates a cross-modal joint module with a specialized multi-stage data pipeline, achieving exceptional audio-visual synchronization and superior generation quality. To ensure practical utility, we implement meticulous post-training optimizations, including Supervised Fine-Tuning (SFT) on high-quality datasets and Reinforcement Learning from Human Feedback (RLHF) with multi-dimensional reward models. Furthermore, we introduce an acceleration framework that boosts inference speed by over 10X. Seedance 1.5 pro distinguishes itself through precise multilingual and dialect lip-syncing, dynamic cinematic camera control, and enhanced narrative coherence, positioning it as a robust engine for professional-grade content creation. Seedance 1.5 pro is now accessible on Volcano Engine at https://console.volcengine.com/ark/region:ark+cn-beijing/experience/vision?type=GenVideo.

Xiaodi Li, Pan Xie, Yi Ren et al.

Audio-driven human animation has attracted wide attention thanks to its practical applications. However, critical challenges remain in generating high-resolution, long-duration videos with consistent appearance and natural hand motions. Existing methods extend videos using overlapping motion frames but suffer from error accumulation, leading to identity drift, color shifts, and scene instability. Additionally, hand movements are poorly modeled, resulting in noticeable distortions and misalignment with the audio. In this work, we propose InfinityHuman, a coarse-to-fine framework that first generates audio-synchronized representations, then progressively refines them into high-resolution, long-duration videos using a pose-guided refiner. Since pose sequences are decoupled from appearance and resist temporal degradation, our pose-guided refiner employs stable poses and the initial frame as a visual anchor to reduce drift and improve lip synchronization. Moreover, to enhance semantic accuracy and gesture realism, we introduce a hand-specific reward mechanism trained with high-quality hand motion data. Experiments on the EMTD and HDTF datasets show that InfinityHuman achieves state-of-the-art performance in video quality, identity preservation, hand accuracy, and lip-sync. Ablation studies further confirm the effectiveness of each module. Code will be made public.

Weifeng Chen, Yatai Ji, Jie Wu et al.

Recent advances in text-to-image (T2I) diffusion models have enabled impressive image generation capabilities guided by text prompts. However, extending these techniques to video generation remains challenging, with existing text-to-video (T2V) methods often struggling to produce high-quality and motion-consistent videos. In this work, we introduce Control-A-Video, a controllable T2V diffusion model that can generate videos conditioned on text prompts and reference control maps like edge and depth maps. To tackle video quality and motion consistency issues, we propose novel strategies to incorporate content prior and motion prior into the diffusion-based generation process. Specifically, we employ a first-frame condition scheme to transfer video generation from the image domain. Additionally, we introduce residual-based and optical flow-based noise initialization to infuse motion priors from reference videos, promoting relevance among frame latents for reduced flickering. Furthermore, we present a Spatio-Temporal Reward Feedback Learning (ST-ReFL) algorithm that optimizes the video diffusion model using multiple reward models for video quality and motion consistency, leading to superior outputs. Comprehensive experiments demonstrate that our framework generates higher-quality, more consistent videos compared to existing state-of-the-art methods in controllable text-to-video generation

Pan Xie, Zexian Li, Xiaohui Hu

Conditional masked language models (CMLM) have shown impressive progress in non-autoregressive machine translation (NAT). They learn the conditional translation model by predicting the random masked subset in the target sentence. Based on the CMLM framework, we introduce Multi-view Subset Regularization (MvSR), a novel regularization method to improve the performance of the NAT model. Specifically, MvSR consists of two parts: (1) \textit{shared mask consistency}: we forward the same target with different mask strategies, and encourage the predictions of shared mask positions to be consistent with each other. (2) \textit{model consistency}, we maintain an exponential moving average of the model weights, and enforce the predictions to be consistent between the average model and the online model. Without changing the CMLM-based architecture, our approach achieves remarkable performance on three public benchmarks with 0.36-1.14 BLEU gains over previous NAT models. Moreover, compared with the stronger Transformer baseline, we reduce the gap to 0.01-0.44 BLEU scores on small datasets (WMT16 RO$\leftrightarrow$EN and IWSLT DE$\rightarrow$EN).

Pan Xie, Zhi Cui, Yao Du et al.

Continuous sign language recognition (cSLR) is a public significant task that transcribes a sign language video into an ordered gloss sequence. It is important to capture the fine-grained gloss-level details, since there is no explicit alignment between sign video frames and the corresponding glosses. Among the past works, one promising way is to adopt a one-dimensional convolutional network (1D-CNN) to temporally fuse the sequential frames. However, CNNs are agnostic to similarity or dissimilarity, and thus are unable to capture local consistent semantics within temporally neighboring frames. To address the issue, we propose to adaptively fuse local features via temporal similarity for this task. Specifically, we devise a Multi-scale Local-Temporal Similarity Fusion Network (mLTSF-Net) as follows: 1) In terms of a specific video frame, we firstly select its similar neighbours with multi-scale receptive regions to accommodate different lengths of glosses. 2) To ensure temporal consistency, we then use position-aware convolution to temporally convolve each scale of selected frames. 3) To obtain a local-temporally enhanced frame-wise representation, we finally fuse the results of different scales using a content-dependent aggregator. We train our model in an end-to-end fashion, and the experimental results on RWTH-PHOENIX-Weather 2014 datasets (RWTH) demonstrate that our model achieves competitive performance compared with several state-of-the-art models.

Pan Xie, Mengyi Zhao, Xiaohui Hu

Since the superiority of Transformer in learning long-term dependency, the sign language Transformer model achieves remarkable progress in Sign Language Recognition (SLR) and Translation (SLT). However, there are several issues with the Transformer that prevent it from better sign language understanding. The first issue is that the self-attention mechanism learns sign video representation in a frame-wise manner, neglecting the temporal semantic structure of sign gestures. Secondly, the attention mechanism with absolute position encoding is direction and distance unaware, thus limiting its ability. To address these issues, we propose a new model architecture, namely PiSLTRc, with two distinctive characteristics: (i) content-aware and position-aware convolution layers. Specifically, we explicitly select relevant features using a novel content-aware neighborhood gathering method. Then we aggregate these features with position-informed temporal convolution layers, thus generating robust neighborhood-enhanced sign representation. (ii) injecting the relative position information to the attention mechanism in the encoder, decoder, and even encoder-decoder cross attention. Compared with the vanilla Transformer model, our model performs consistently better on three large-scale sign language benchmarks: PHOENIX-2014, PHOENIX-2014-T and CSL. Furthermore, extensive experiments demonstrate that the proposed method achieves state-of-the-art performance on translation quality with $+1.6$ BLEU improvements.

Pan Xie, Zhi Cui, Xiuyin Chen et al.

Non-autoregressive models generate target words in a parallel way, which achieve a faster decoding speed but at the sacrifice of translation accuracy. To remedy a flawed translation by non-autoregressive models, a promising approach is to train a conditional masked translation model (CMTM), and refine the generated results within several iterations. Unfortunately, such approach hardly considers the \textit{sequential dependency} among target words, which inevitably results in a translation degradation. Hence, instead of solely training a Transformer-based CMTM, we propose a Self-Review Mechanism to infuse sequential information into it. Concretely, we insert a left-to-right mask to the same decoder of CMTM, and then induce it to autoregressively review whether each generated word from CMTM is supposed to be replaced or kept. The experimental results (WMT14 En$\leftrightarrow$De and WMT16 En$\leftrightarrow$Ro) demonstrate that our model uses dramatically less training computations than the typical CMTM, as well as outperforms several state-of-the-art non-autoregressive models by over 1 BLEU. Through knowledge distillation, our model even surpasses a typical left-to-right Transformer model, while significantly speeding up decoding.