Yuxuan Wang

Ziyang Ma, Yakun Song, Chenpeng Du et al.

Dialogue serves as the most natural manner of human-computer interaction (HCI). Recent advancements in speech language models (SLM) have significantly enhanced speech-based conversational AI. However, these models are limited to turn-based conversation, lacking the ability to interact with humans in real-time spoken scenarios, for example, being interrupted when the generated content is not satisfactory. To address these limitations, we explore full duplex modeling (FDM) in interactive speech language models (iSLM), focusing on enhancing real-time interaction and, more explicitly, exploring the quintessential ability of interruption. We introduce a novel model design, namely listening-while-speaking language model (LSLM), an end-to-end system equipped with both listening and speaking channels. Our LSLM employs a token-based decoder-only TTS for speech generation and a streaming self-supervised learning (SSL) encoder for real-time audio input. LSLM fuses both channels for autoregressive generation and detects turn-taking in real time. Three fusion strategies -- early fusion, middle fusion, and late fusion -- are explored, with middle fusion achieving an optimal balance between speech generation and real-time interaction. Two experimental settings, command-based FDM and voice-based FDM, demonstrate LSLM's robustness to noise and sensitivity to diverse instructions. Our results highlight LSLM's capability to achieve duplex communication with minimal impact on existing systems. This study aims to advance the development of interactive speech dialogue systems, enhancing their applicability in real-world contexts.

Yuxuan Wang, Haixu Wu, Jiaxiang Dong et al.

Deep models have demonstrated remarkable performance in time series forecasting. However, due to the partially-observed nature of real-world applications, solely focusing on the target of interest, so-called endogenous variables, is usually insufficient to guarantee accurate forecasting. Notably, a system is often recorded into multiple variables, where the exogenous variables can provide valuable external information for endogenous variables. Thus, unlike well-established multivariate or univariate forecasting paradigms that either treat all the variables equally or ignore exogenous information, this paper focuses on a more practical setting: time series forecasting with exogenous variables. We propose a novel approach, TimeXer, to ingest external information to enhance the forecasting of endogenous variables. With deftly designed embedding layers, TimeXer empowers the canonical Transformer with the ability to reconcile endogenous and exogenous information, where patch-wise self-attention and variate-wise cross-attention are used simultaneously. Moreover, global endogenous tokens are learned to effectively bridge the causal information underlying exogenous series into endogenous temporal patches. Experimentally, TimeXer achieves consistent state-of-the-art performance on twelve real-world forecasting benchmarks and exhibits notable generality and scalability. Code is available at this repository: https://github.com/thuml/TimeXer.

Yuxuan Wang, Yiqi Song, Cihang Xie et al.

Recent advancements in large-scale video-language models have shown significant potential for real-time planning and detailed interactions. However, their high computational demands and the scarcity of annotated datasets limit their practicality for academic researchers. In this work, we introduce VideoLLaMB, a novel and efficient framework for long video understanding that leverages recurrent memory bridges and temporal memory tokens to enable seamless encoding of entire video sequences with preserved semantic continuity. Central to our approach is a SceneTiling algorithm that segments videos into coherent semantic units, facilitating robust understanding across tasks without requiring additional training. VideoLLaMB achieves state-of-the-art performance, surpassing existing models by 4.2 points on four VideoQA benchmarks and by 2.06 points on egocentric planning tasks. Notably, it maintains strong performance under extreme video length scaling (up to 8 times) and excels at fine-grained frame retrieval on our proposed Needle in a Video Haystack (NIAVH) benchmark. With linear GPU memory scaling, VideoLLaMB processes up to 320 frames using a single Nvidia A100 GPU, despite being trained on only 16 frames-offering an unprecedented balance of accuracy, scalability, and cost-effectiveness. This makes it highly accessible and practical for the academic community.

Yuxuan Wang, Xuanyu Yi, Qingshan Xu et al.

Personalizing 3D scenes from a single reference image enables intuitive user-guided editing, which requires achieving both multi-view consistency across perspectives and referential consistency with the input image. However, these goals are particularly challenging due to the viewpoint bias caused by the limited perspective provided in a single image. Lacking the mechanisms to effectively expand reference information beyond the original view, existing methods of image-conditioned 3DGS personalization often suffer from this viewpoint bias and struggle to produce consistent results. Therefore, in this paper, we present Consistent Personalization for 3D Gaussian Splatting (CP-GS), a framework that progressively propagates the single-view reference appearance to novel perspectives. In particular, CP-GS integrates pre-trained image-to-3D generation and iterative LoRA fine-tuning to extract and extend the reference appearance, and finally produces faithful multi-view guidance images and the personalized 3DGS outputs through a view-consistent generation process guided by geometric cues. Extensive experiments on real-world scenes show that our CP-GS effectively mitigates the viewpoint bias, achieving high-quality personalization that significantly outperforms existing methods. The code will be released at https://github.com/Yuxuan-W/CP-GS.

Yuxuan Wang, Xuanyu Yi, Zike Wu et al.

The advent of 3D Gaussian Splatting (3DGS) has revolutionized 3D editing, offering efficient, high-fidelity rendering and enabling precise local manipulations. Currently, diffusion-based 2D editing models are harnessed to modify multi-view rendered images, which then guide the editing of 3DGS models. However, this approach faces a critical issue of multi-view inconsistency, where the guidance images exhibit significant discrepancies across views, leading to mode collapse and visual artifacts of 3DGS. To this end, we introduce View-consistent Editing (VcEdit), a novel framework that seamlessly incorporates 3DGS into image editing processes, ensuring multi-view consistency in edited guidance images and effectively mitigating mode collapse issues. VcEdit employs two innovative consistency modules: the Cross-attention Consistency Module and the Editing Consistency Module, both designed to reduce inconsistencies in edited images. By incorporating these consistency modules into an iterative pattern, VcEdit proficiently resolves the issue of multi-view inconsistency, facilitating high-quality 3DGS editing across a diverse range of scenes. Further video results are shown in http://vcedit.github.io.

Qingshan Xu, Jiequan Cui, Xuanyu Yi et al.

3D Gaussian Splatting (3DGS) has demonstrated impressive Novel View Synthesis (NVS) results in a real-time rendering manner. During training, it relies heavily on the average magnitude of view-space positional gradients to grow Gaussians to reduce rendering loss. However, this average operation smooths the positional gradients from different viewpoints and rendering errors from different pixels, hindering the growth and optimization of many defective Gaussians. This leads to strong spurious artifacts in some areas. To address this problem, we propose Hard Gaussian Splatting, dubbed HGS, which considers multi-view significant positional gradients and rendering errors to grow hard Gaussians that fill the gaps of classical Gaussian Splatting on 3D scenes, thus achieving superior NVS results. In detail, we present positional gradient driven HGS, which leverages multi-view significant positional gradients to uncover hard Gaussians. Moreover, we propose rendering error guided HGS, which identifies noticeable pixel rendering errors and potentially over-large Gaussians to jointly mine hard Gaussians. By growing and optimizing these hard Gaussians, our method helps to resolve blurring and needle-like artifacts. Experiments on various datasets demonstrate that our method achieves state-of-the-art rendering quality while maintaining real-time efficiency.

Yuan Zhou, Junbao Zhou, Qingshan Xu et al.

Drag-Based Image Editing (DBIE), which allows users to manipulate images by directly dragging objects within them, has recently attracted much attention from the community. However, it faces two key challenges: (\emph{\textcolor{magenta}{i}}) point-based drag is often highly ambiguous and difficult to align with users' intentions; (\emph{\textcolor{magenta}{ii}}) current DBIE methods primarily rely on alternating between motion supervision and point tracking, which is not only cumbersome but also fails to produce high-quality results. These limitations motivate us to explore DBIE from a new perspective -- redefining it as deformation, rotation, and translation of user-specified handle regions. Thereby, by requiring users to explicitly specify both drag areas and types, we can effectively address the ambiguity issue. Furthermore, we propose a simple-yet-effective editing framework, dubbed \textcolor{SkyBlue}{\textbf{DragNeXt}}. It unifies DBIE as a Latent Region Optimization (LRO) problem and solves it through Progressive Backward Self-Intervention (PBSI), simplifying the overall procedure of DBIE while further enhancing quality by fully leveraging region-level structure information and progressive guidance from intermediate drag states. We validate \textcolor{SkyBlue}{\textbf{DragNeXt}} on our NextBench, and extensive experiments demonstrate that our proposed method can significantly outperform existing approaches. Code will be released on github.

Ziyang Ma, Xiquan Li, Yakun Song et al.

Recent advancements in large audio language models (LALMs) have demonstrated impressive results and promising prospects in universal understanding and reasoning across speech, music, and general sound. However, these models still lack the ability to recognize their knowledge boundaries and refuse to answer questions they don't know proactively. While there have been successful attempts to enhance the reliability of LLMs, reliable LALMs remain largely unexplored. In this paper, we systematically investigate various approaches towards reliable LALMs, including training-free methods such as multi-modal chain-of-thought (MCoT), and training-based methods such as supervised fine-tuning (SFT). Besides, we identify the limitations of previous evaluation metrics and propose a new metric, the Reliability Gain Index (RGI), to assess the effectiveness of different reliable methods. Our findings suggest that both training-free and training-based methods enhance the reliability of LALMs to different extents. Moreover, we find that awareness of reliability is a "meta ability", which can be transferred across different audio modalities, although significant structural and content differences exist among sound, music, and speech.

Kequan Chen, Pan Liu, Yuxuan Wang et al.

Accurate prediction of traffic crash risks for individual vehicles is essential for enhancing vehicle safety. While significant attention has been given to traffic crash risk prediction, existing studies face two main challenges: First, due to the scarcity of individual vehicle data before crashes, most models rely on hypothetical scenarios deemed dangerous by researchers. This raises doubts about their applicability to actual pre-crash conditions. Second, some crash risk prediction frameworks were learned from dashcam videos. Although such videos capture the pre-crash behavior of individual vehicles, they often lack critical information about the movements of surrounding vehicles. However, the interaction between a vehicle and its surrounding vehicles is highly influential in crash occurrences. To overcome these challenges, we propose a novel non-stationary extreme value theory (EVT), where the covariate function is optimized in a nonlinear fashion using a graph attention network. The EVT component incorporates the stochastic nature of crashes through probability distribution, which enhances model interpretability. Notably, the nonlinear covariate function enables the model to capture the interactive behavior between the target vehicle and its multiple surrounding vehicles, facilitating crash risk prediction across different driving tasks. We train and test our model using 100 sets of vehicle trajectory data before real crashes, collected via drones over three years from merging and weaving segments. We demonstrate that our model successfully learns micro-level precursors of crashes and fits a more accurate distribution with the aid of the nonlinear covariate function. Our experiments on the testing dataset show that the proposed model outperforms existing models by providing more accurate predictions for both rear-end and sideswipe crashes simultaneously.

Yongxin Zhu, Jiawei Chen, Yuanzhe Chen et al.

We introduce Heptapod, an image autoregressive model that adheres to the foundational principles of language modeling. Heptapod employs \textbf{causal attention}, \textbf{eliminates reliance on CFG}, and \textbf{eschews the trend of semantic tokenizers}. Our key innovation is \textit{next 2D distribution prediction}: a causal Transformer with reconstruction-focused visual tokenizer, learns to predict the distribution over the entire 2D spatial grid of images at each timestep. This learning objective unifies the sequential modeling of autoregressive framework with the holistic self-supervised learning of masked autoencoding, enabling the model to capture comprehensive image semantics via generative training. On the ImageNet generation benchmark, Heptapod achieves an FID of $2.70$, significantly outperforming previous causal autoregressive approaches. We hope our work inspires a principled rethinking of language modeling on visual signals and beyond.