Bohan Li

Yihan Wu, Xu Tan, Bohan Li et al. · microsoft-research

Adaptive text to speech (TTS) can synthesize new voices in zero-shot scenarios efficiently, by using a well-trained source TTS model without adapting it on the speech data of new speakers. Considering seen and unseen speakers have diverse characteristics, zero-shot adaptive TTS requires strong generalization ability on speaker characteristics, which brings modeling challenges. In this paper, we develop AdaSpeech 4, a zero-shot adaptive TTS system for high-quality speech synthesis. We model the speaker characteristics systematically to improve the generalization on new speakers. Generally, the modeling of speaker characteristics can be categorized into three steps: extracting speaker representation, taking this speaker representation as condition, and synthesizing speech/mel-spectrogram given this speaker representation. Accordingly, we improve the modeling in three steps: 1) To extract speaker representation with better generalization, we factorize the speaker characteristics into basis vectors and extract speaker representation by weighted combining of these basis vectors through attention. 2) We leverage conditional layer normalization to integrate the extracted speaker representation to TTS model. 3) We propose a novel supervision loss based on the distribution of basis vectors to maintain the corresponding speaker characteristics in generated mel-spectrograms. Without any fine-tuning, AdaSpeech 4 achieves better voice quality and similarity than baselines in multiple datasets.

Hongrui Xuan, Yi Liu, Bohan Li et al. · cmu

A well-designed recommender system can accurately capture the attributes of users and items, reflecting the unique preferences of individuals. Traditional recommendation techniques usually focus on modeling the singular type of behaviors between users and items. However, in many practical recommendation scenarios (e.g., social media, e-commerce), there exist multi-typed interactive behaviors in user-item relationships, such as click, tag-as-favorite, and purchase in online shopping platforms. Thus, how to make full use of multi-behavior information for recommendation is of great importance to the existing system, which presents challenges in two aspects that need to be explored: (1) Utilizing users' personalized preferences to capture multi-behavioral dependencies; (2) Dealing with the insufficient recommendation caused by sparse supervision signal for target behavior. In this work, we propose a Knowledge Enhancement Multi-Behavior Contrastive Learning Recommendation (KMCLR) framework, including two Contrastive Learning tasks and three functional modules to tackle the above challenges, respectively. In particular, we design the multi-behavior learning module to extract users' personalized behavior information for user-embedding enhancement, and utilize knowledge graph in the knowledge enhancement module to derive more robust knowledge-aware representations for items. In addition, in the optimization stage, we model the coarse-grained commonalities and the fine-grained differences between multi-behavior of users to further improve the recommendation effect. Extensive experiments and ablation tests on the three real-world datasets indicate our KMCLR outperforms various state-of-the-art recommendation methods and verify the effectiveness of our method.

Yutai Hou, Hongyuan Dong, Xinghao Wang et al. · cmu

Prompting method is regarded as one of the crucial progress for few-shot nature language processing. Recent research on prompting moves from discrete tokens based ``hard prompts'' to continuous ``soft prompts'', which employ learnable vectors as pseudo prompt tokens and achieve better performance. Though showing promising prospects, these soft-prompting methods are observed to rely heavily on good initialization to take effect. Unfortunately, obtaining a perfect initialization for soft prompts requires understanding of inner language models working and elaborate design, which is no easy task and has to restart from scratch for each new task. To remedy this, we propose a generalized soft prompting method called MetaPrompting, which adopts the well-recognized model-agnostic meta-learning algorithm to automatically find better prompt initialization that facilitates fast adaptation to new prompting tasks.Extensive experiments show MetaPrompting tackles soft prompt initialization problem and brings significant improvement on four different datasets (over 6 points improvement in accuracy for 1-shot setting), achieving new state-of-the-art performance.

Bohan Li, Yasheng Sun, Zhujin Liang et al.

3D semantic scene completion (SSC) is an ill-posed perception task that requires inferring a dense 3D scene from limited observations. Previous camera-based methods struggle to predict accurate semantic scenes due to inherent geometric ambiguity and incomplete observations. In this paper, we resort to stereo matching technique and bird's-eye-view (BEV) representation learning to address such issues in SSC. Complementary to each other, stereo matching mitigates geometric ambiguity with epipolar constraint while BEV representation enhances the hallucination ability for invisible regions with global semantic context. However, due to the inherent representation gap between stereo geometry and BEV features, it is non-trivial to bridge them for dense prediction task of SSC. Therefore, we further develop a unified occupancy-based framework dubbed BRGScene, which effectively bridges these two representations with dense 3D volumes for reliable semantic scene completion. Specifically, we design a novel Mutual Interactive Ensemble (MIE) block for pixel-level reliable aggregation of stereo geometry and BEV features. Within the MIE block, a Bi-directional Reliable Interaction (BRI) module, enhanced with confidence re-weighting, is employed to encourage fine-grained interaction through mutual guidance. Besides, a Dual Volume Ensemble (DVE) module is introduced to facilitate complementary aggregation through channel-wise recalibration and multi-group voting. Our method outperforms all published camera-based methods on SemanticKITTI for semantic scene completion. Our code is available on https://github.com/Arlo0o/StereoScene.

Yihan Wu, Junliang Guo, Xu Tan et al. · microsoft-research

Video dubbing aims to translate the original speech in a film or television program into the speech in a target language, which can be achieved with a cascaded system consisting of speech recognition, machine translation and speech synthesis. To ensure the translated speech to be well aligned with the corresponding video, the length/duration of the translated speech should be as close as possible to that of the original speech, which requires strict length control. Previous works usually control the number of words or characters generated by the machine translation model to be similar to the source sentence, without considering the isochronicity of speech as the speech duration of words/characters in different languages varies. In this paper, we propose a machine translation system tailored for the task of video dubbing, which directly considers the speech duration of each token in translation, to match the length of source and target speech. Specifically, we control the speech length of generated sentence by guiding the prediction of each word with the duration information, including the speech duration of itself as well as how much duration is left for the remaining words. We design experiments on four language directions (German -> English, Spanish -> English, Chinese <-> English), and the results show that the proposed method achieves better length control ability on the generated speech than baseline methods. To make up the lack of real-world datasets, we also construct a real-world test set collected from films to provide comprehensive evaluations on the video dubbing task.

Bohan Li, Ye Yuan, Dingkang Liang et al.

Recently, most handwritten mathematical expression recognition (HMER) methods adopt the encoder-decoder networks, which directly predict the markup sequences from formula images with the attention mechanism. However, such methods may fail to accurately read formulas with complicated structure or generate long markup sequences, as the attention results are often inaccurate due to the large variance of writing styles or spatial layouts. To alleviate this problem, we propose an unconventional network for HMER named Counting-Aware Network (CAN), which jointly optimizes two tasks: HMER and symbol counting. Specifically, we design a weakly-supervised counting module that can predict the number of each symbol class without the symbol-level position annotations, and then plug it into a typical attention-based encoder-decoder model for HMER. Experiments on the benchmark datasets for HMER validate that both joint optimization and counting results are beneficial for correcting the prediction errors of encoder-decoder models, and CAN consistently outperforms the state-of-the-art methods. In particular, compared with an encoder-decoder model for HMER, the extra time cost caused by the proposed counting module is marginal. The source code is available at https://github.com/LBH1024/CAN.

Bohan Li, Shi Lian, Hankun Wang et al.

Unified speech foundation models require a holistic tokenization space that is both learnable by language models and decodable into high-quality waveforms. Existing speech tokenizers, however, often fail to satisfy these requirements simultaneously, leading to increased architectural complexity and more involved training designs. We propose HoliTok, a continuous Holistic speech Tokenization model designed for unified generation-understanding modeling. HoliTok encodes 48~kHz speech into a compact 25~Hz sequence of 128-dimensional latents. It is trained with a progressive strategy that jointly preserves signal-level fidelity, incorporates semantic information, and maintains strong latent learnability. Based on this tokenization, we build a unified AR+DiT model for speech synthesis and recognition, where the same latent sequence supports both generation-specific and unified generation-understanding tasks. Experiments show that HoliTok achieves competitive reconstruction fidelity, improves generative learnability for high-quality and controllable synthesis, and, among the evaluated representations, is the only one that operates robustly in our unified generation-understanding architecture without additional optimization tricks. These results suggest that HoliTok serves as an effective speech tokenizer and a foundational representation interface for unified spoken language modeling. The code is available at: https://github.com/bovod-sjtu/HoliTok.

Zhihan Li, Hankun Wang, Yiwei Guo et al.

Automatic speech recognition systems commonly rely on reference transcriptions for evaluation, while reference-free approaches often depend on internal confidence estimation or auxiliary language models. We propose READ (Reference-free Hypothesis Evaluation with Acoustic Discrepancy), a novel metric that evaluates ASR hypotheses directly from the speech signal. READ emphasizes the acoustic grounding of hypotheses. It uses a pretrained auto-regressive TTS model to compute the conditional likelihood of speech tokens given a text hypothesis, to measure fine-grained acoustic discrepancy between speech and text. Without additional training, READ can be applied for hypothesis refinement. Experiments show that READ correlates with specific recognition errors and improves ASR outputs, achieving up to 20\% relative error rate reduction, with particularly strong gains under noisy conditions.

Lianying Yin, Yijun Wang, Tianyu He et al. · microsoft-research

Although previous co-speech gesture generation methods are able to synthesize motions in line with speech content, it is still not enough to handle diverse and complicated motion distribution. The key challenges are: 1) the one-to-many nature between the speech content and gestures; 2) the correlation modeling between the body joints. In this paper, we present a novel framework (EMoG) to tackle the above challenges with denoising diffusion models: 1) To alleviate the one-to-many problem, we incorporate emotion clues to guide the generation process, making the generation much easier; 2) To model joint correlation, we propose to decompose the difficult gesture generation into two sub-problems: joint correlation modeling and temporal dynamics modeling. Then, the two sub-problems are explicitly tackled with our proposed Joint Correlation-aware transFormer (JCFormer). Through extensive evaluations, we demonstrate that our proposed method surpasses previous state-of-the-art approaches, offering substantial superiority in gesture synthesis.

Yutai Hou, Cheng Chen, Xianzhen Luo et al.

Prompting methods recently achieve impressive success in few-shot learning. These methods modify input samples with prompt sentence pieces, and decode label tokens to map samples to corresponding labels. However, such a paradigm is very inefficient for the task of slot tagging. Since slot tagging samples are multiple consecutive words in a sentence, the prompting methods have to enumerate all n-grams token spans to find all the possible slots, which greatly slows down the prediction. To tackle this, we introduce an inverse paradigm for prompting. Different from the classic prompts mapping tokens to labels, we reversely predict slot values given slot types. Such inverse prompting only requires a one-turn prediction for each slot type and greatly speeds up the prediction. Besides, we propose a novel Iterative Prediction Strategy, from which the model learns to refine predictions by considering the relations between different slot types. We find, somewhat surprisingly, the proposed method not only predicts faster but also significantly improves the effect (improve over 6.1 F1-scores on 10-shot setting) and achieves new state-of-the-art performance.

Wenwen Yu, Chengquan Zhang, Haoyu Cao et al.

Structured text extraction is one of the most valuable and challenging application directions in the field of Document AI. However, the scenarios of past benchmarks are limited, and the corresponding evaluation protocols usually focus on the submodules of the structured text extraction scheme. In order to eliminate these problems, we organized the ICDAR 2023 competition on Structured text extraction from Visually-Rich Document images (SVRD). We set up two tracks for SVRD including Track 1: HUST-CELL and Track 2: Baidu-FEST, where HUST-CELL aims to evaluate the end-to-end performance of Complex Entity Linking and Labeling, and Baidu-FEST focuses on evaluating the performance and generalization of Zero-shot / Few-shot Structured Text extraction from an end-to-end perspective. Compared to the current document benchmarks, our two tracks of competition benchmark enriches the scenarios greatly and contains more than 50 types of visually-rich document images (mainly from the actual enterprise applications). The competition opened on 30th December, 2022 and closed on 24th March, 2023. There are 35 participants and 91 valid submissions received for Track 1, and 15 participants and 26 valid submissions received for Track 2. In this report we will presents the motivation, competition datasets, task definition, evaluation protocol, and submission summaries. According to the performance of the submissions, we believe there is still a large gap on the expected information extraction performance for complex and zero-shot scenarios. It is hoped that this competition will attract many researchers in the field of CV and NLP, and bring some new thoughts to the field of Document AI.

Bohan Li, Jiajun Deng, Wenyao Zhang et al.

Camera-based 3D semantic scene completion (SSC) is pivotal for predicting complicated 3D layouts with limited 2D image observations. The existing mainstream solutions generally leverage temporal information by roughly stacking history frames to supplement the current frame, such straightforward temporal modeling inevitably diminishes valid clues and increases learning difficulty. To address this problem, we present HTCL, a novel Hierarchical Temporal Context Learning paradigm for improving camera-based semantic scene completion. The primary innovation of this work involves decomposing temporal context learning into two hierarchical steps: (a) cross-frame affinity measurement and (b) affinity-based dynamic refinement. Firstly, to separate critical relevant context from redundant information, we introduce the pattern affinity with scale-aware isolation and multiple independent learners for fine-grained contextual correspondence modeling. Subsequently, to dynamically compensate for incomplete observations, we adaptively refine the feature sampling locations based on initially identified locations with high affinity and their neighboring relevant regions. Our method ranks $1^{st}$ on the SemanticKITTI benchmark and even surpasses LiDAR-based methods in terms of mIoU on the OpenOccupancy benchmark. Our code is available on https://github.com/Arlo0o/HTCL.

Bohan Li, Xiao Xu, Xinghao Wang et al.

Existing image augmentation methods consist of two categories: perturbation-based methods and generative methods. Perturbation-based methods apply pre-defined perturbations to augment an original image, but only locally vary the image, thus lacking image diversity. In contrast, generative methods bring more image diversity in the augmented images but may not preserve semantic consistency, thus incorrectly changing the essential semantics of the original image. To balance image diversity and semantic consistency in augmented images, we propose SGID, a Semantic-guided Generative Image augmentation method with Diffusion models for image classification. Specifically, SGID employs diffusion models to generate augmented images with good image diversity. More importantly, SGID takes image labels and captions as guidance to maintain semantic consistency between the augmented and original images. Experimental results show that SGID outperforms the best augmentation baseline by 1.72% on ResNet-50 (from scratch), 0.33% on ViT (ImageNet-21k), and 0.14% on CLIP-ViT (LAION-2B). Moreover, SGID can be combined with other image augmentation baselines and further improves the overall performance. We demonstrate the semantic consistency and image diversity of SGID through quantitative human and automated evaluations, as well as qualitative case studies.

Baao Xie, Bohan Li, Zequn Zhang et al.

3D representation disentanglement aims to identify, decompose, and manipulate the underlying explanatory factors of 3D data, which helps AI fundamentally understand our 3D world. This task is currently under-explored and poses great challenges: (i) the 3D representations are complex and in general contains much more information than 2D image; (ii) many 3D representations are not well suited for gradient-based optimization, let alone disentanglement. To address these challenges, we use NeRF as a differentiable 3D representation, and introduce a self-supervised Navigation to identify interpretable semantic directions in the latent space. To our best knowledge, this novel method, dubbed NaviNeRF, is the first work to achieve fine-grained 3D disentanglement without any priors or supervisions. Specifically, NaviNeRF is built upon the generative NeRF pipeline, and equipped with an Outer Navigation Branch and an Inner Refinement Branch. They are complementary -- the outer navigation is to identify global-view semantic directions, and the inner refinement dedicates to fine-grained attributes. A synergistic loss is further devised to coordinate two branches. Extensive experiments demonstrate that NaviNeRF has a superior fine-grained 3D disentanglement ability than the previous 3D-aware models. Its performance is also comparable to editing-oriented models relying on semantic or geometry priors.

Bohan Li, Longxu Dou, Yutai Hou et al.

Prompt-based learning has shown considerable promise in reformulating various downstream tasks as cloze problems by combining original input with a predetermined template. This approach demonstrates its effectiveness, especially in few-shot learning scenarios, where the model is trained on a scarce amount of data. Despite its successes, the limited templates and text in few-shot prompt-based learning scenarios leave significant room for performance improvement. Moreover, existing methods sometimes resort to model ensembles, which, while effective, could potentially hamper model efficiency due to increased computational demands. To address these issues, we introduce MixPro, an augmentation method designed to augment both the vanilla input text and the templates. We implement this through the token-level, the sentence-level, and the template-level Mixup strategies. The experimental results on five few-shot datasets show that MixPro outperforms other augmentation baselines, improving model performance by an average of 5.08% compared to before augmentation.

Bowen Deng, Bohan Li, Matthew Cox et al.

Computational materials science and chemistry span vast knowledge domains and fractured software ecosystems. Although large language models (LLMs) have demonstrated research capabilities, scaling monolithic agents to manage the rigor and complexity of atomistic research remains a challenge. Here, we introduce AtomisticSkills, an open-source harness framework that empowers general-purpose AI coding agents to conduct atomistic research across materials science, chemistry, and drug discovery. By hierarchically decomposing scientific workflows into agent skills and tools, AtomisticSkills provides agents with modular, extensible, and plug-and-play research capabilities. The framework integrates more than 100 human-curated multidisciplinary skills, including database access, thermodynamics and kinetics modeling, and diverse simulation engines employing machine learning interatomic potentials (MLIPs) and density functional theory (DFT). We validate its functional coverage against scientific literature and demonstrate robust orchestration capabilities across diverse scientific campaigns: generative design of Li-ion solid-state electrolytes, high-throughput screening of metal-organic frameworks for CO2 capture, autonomous MLIP benchmarking and fine-tuning, multi-stage structure-based virtual screening for drug design, multimodal X-ray diffraction pattern analysis, and screening of Fe-oxide catalysts for oxygen evolution reaction. AtomisticSkills provides a critical agent infrastructure towards building fully autonomous AI scientists.

Yunlong Feng, Bohan Li, Libo Qin et al.

Cross-domain text classification aims to adapt models to a target domain that lacks labeled data. It leverages or reuses rich labeled data from the different but related source domain(s) and unlabeled data from the target domain. To this end, previous work focuses on either extracting domain-invariant features or task-agnostic features, ignoring domain-aware features that may be present in the target domain and could be useful for the downstream task. In this paper, we propose a two-stage framework for cross-domain text classification. In the first stage, we finetune the model with mask language modeling (MLM) and labeled data from the source domain. In the second stage, we further fine-tune the model with self-supervised distillation (SSD) and unlabeled data from the target domain. We evaluate its performance on a public cross-domain text classification benchmark and the experiment results show that our method achieves new state-of-the-art results for both single-source domain adaptations (94.17% $\uparrow$1.03%) and multi-source domain adaptations (95.09% $\uparrow$1.34%).

Hongyang Li, Hao Zhang, Shilong Liu et al.

In this paper, we present TAPTRv2, a Transformer-based approach built upon TAPTR for solving the Tracking Any Point (TAP) task. TAPTR borrows designs from DEtection TRansformer (DETR) and formulates each tracking point as a point query, making it possible to leverage well-studied operations in DETR-like algorithms. TAPTRv2 improves TAPTR by addressing a critical issue regarding its reliance on cost-volume,which contaminates the point queryś content feature and negatively impacts both visibility prediction and cost-volume computation. In TAPTRv2, we propose a novel attention-based position update (APU) operation and use key-aware deformable attention to realize. For each query, this operation uses key-aware attention weights to combine their corresponding deformable sampling positions to predict a new query position. This design is based on the observation that local attention is essentially the same as cost-volume, both of which are computed by dot-production between a query and its surrounding features. By introducing this new operation, TAPTRv2 not only removes the extra burden of cost-volume computation, but also leads to a substantial performance improvement. TAPTRv2 surpasses TAPTR and achieves state-of-the-art performance on many challenging datasets, demonstrating the superiority

Bohan Li, Yasheng Sun, Jingxin Dong et al.

Numerous studies have investigated the pivotal role of reliable 3D volume representation in scene perception tasks, such as multi-view stereo (MVS) and semantic scene completion (SSC). They typically construct 3D probability volumes directly with geometric correspondence, attempting to fully address the scene perception tasks in a single forward pass. However, such a single-step solution makes it hard to learn accurate and convincing volumetric probability, especially in challenging regions like unexpected occlusions and complicated light reflections. Therefore, this paper proposes to decompose the complicated 3D volume representation learning into a sequence of generative steps to facilitate fine and reliable scene perception. Considering the recent advances achieved by strong generative diffusion models, we introduce a multi-step learning framework, dubbed as VPD, dedicated to progressively refining the Volumetric Probability in a Diffusion process. Extensive experiments are conducted on scene perception tasks including multi-view stereo (MVS) and semantic scene completion (SSC), to validate the efficacy of our method in learning reliable volumetric representations. Notably, for the SSC task, our work stands out as the first to surpass LiDAR-based methods on the SemanticKITTI dataset.

Bohan Li, Feiyu Shen, Yiwei Guo et al.

Discretizing speech into tokens and generating them by a decoder-only model have been a promising direction for text-to-speech (TTS) and spoken language modeling (SLM). To shorten the sequence length of speech tokens, acoustic byte-pair encoding (BPE) has emerged in SLM that treats speech tokens from self-supervised semantic representations as characters to further compress the token sequence. But the gain in TTS has not been fully investigated, and the proper choice of acoustic BPE remains unclear. In this work, we conduct a comprehensive study on various settings of acoustic BPE to explore its effectiveness in decoder-only TTS models with semantic speech tokens. Experiments on LibriTTS verify that acoustic BPE uniformly increases the intelligibility and diversity of synthesized speech, while showing different features across BPE settings. Hence, acoustic BPE is a favorable tool for decoder-only TTS.

GigaBrain Team, Boyuan Wang, Bohan Li et al.

Vision-language-action (VLA) models that directly predict multi-step action chunks from current observations face inherent limitations due to constrained scene understanding and weak future anticipation capabilities. In contrast, video world models pre-trained on web-scale video corpora exhibit robust spatiotemporal reasoning and accurate future prediction, making them a natural foundation for enhancing VLA learning. Therefore, we propose \textit{GigaBrain-0.5M*}, a VLA model trained via world model-based reinforcement learning. Built upon \textit{GigaBrain-0.5}, which is pre-trained on over 10,000 hours of robotic manipulation data, whose intermediate version currently ranks first on the international RoboChallenge benchmark. \textit{GigaBrain-0.5M*} further integrates world model-based reinforcement learning via \textit{RAMP} (Reinforcement leArning via world Model-conditioned Policy) to enable robust cross-task adaptation. Empirical results demonstrate that \textit{RAMP} achieves substantial performance gains over the RECAP baseline, yielding improvements of approximately 30\% on challenging tasks including \texttt{Laundry Folding}, \texttt{Box Packing}, and \texttt{Espresso Preparation}. Critically, \textit{GigaBrain-0.5M$^*$} exhibits reliable long-horizon execution, consistently accomplishing complex manipulation tasks without failure as validated by real-world deployment videos on our \href{https://gigabrain05m.github.io}{project page}.

Tianqi Liu, Zhaoxi Chen, Zihao Huang et al.

Recent advances in illumination control extend image-based methods to video, yet still facing a trade-off between lighting fidelity and temporal consistency. Moving beyond relighting, a key step toward generative modeling of real-world scenes is the joint control of camera trajectory and illumination, since visual dynamics are inherently shaped by both geometry and lighting. To this end, we present Light-X, a video generation framework that enables controllable rendering from monocular videos with both viewpoint and illumination control. 1) We propose a disentangled design that decouples geometry and lighting signals: geometry and motion are captured via dynamic point clouds projected along user-defined camera trajectories, while illumination cues are provided by a relit frame consistently projected into the same geometry. These explicit, fine-grained cues enable effective disentanglement and guide high-quality illumination. 2) To address the lack of paired multi-view and multi-illumination videos, we introduce Light-Syn, a degradation-based pipeline with inverse-mapping that synthesizes training pairs from in-the-wild monocular footage. This strategy yields a dataset covering static, dynamic, and AI-generated scenes, ensuring robust training. Extensive experiments show that Light-X outperforms baseline methods in joint camera-illumination control and surpasses prior video relighting methods under both text- and background-conditioned settings.

Yunfan Gao, Yun Xiong, Wenlong Wu et al.

Recent advancements in Large Language Models (LLMs) have expanded their context windows to unprecedented lengths, sparking debates about the necessity of Retrieval-Augmented Generation (RAG). To address the fragmented evaluation paradigms and limited cases in existing Needle-in-a-Haystack (NIAH), this paper introduces U-NIAH, a unified framework that systematically compares LLMs and RAG methods in controlled long context settings. Our framework extends beyond traditional NIAH by incorporating multi-needle, long-needle, and needle-in-needle configurations, along with different retrieval settings, while leveraging the synthetic Starlight Academy dataset-a fictional magical universe-to eliminate biases from pre-trained knowledge. Through extensive experiments, we investigate three research questions: (1) performance trade-offs between LLMs and RAG, (2) error patterns in RAG, and (3) RAG's limitations in complex settings. Our findings show that RAG significantly enhances smaller LLMs by mitigating the "lost-in-the-middle" effect and improving robustness, achieving an 82.58% win-rate over LLMs. However, we observe that retrieval noise and reverse chunk ordering degrade performance, while surprisingly, advanced reasoning LLMs exhibit reduced RAG compatibility due to sensitivity to semantic distractors. We identify typical error patterns including omission due to noise, hallucination under high noise critical condition, and self-doubt behaviors. Our work not only highlights the complementary roles of RAG and LLMs, but also provides actionable insights for optimizing deployments. Code: https://github.com/Tongji-KGLLM/U-NIAH.

Hao Tang, Songyun Xie, Xinzhou Xie et al.

Traditional video-induced physiological datasets usually rely on whole-trial labels, which introduce temporal label noise in dynamic emotion recognition. We present FIRMED, a peak-centered multimodal dataset based on an immediate-recall annotation paradigm, with synchronized EEG, ECG, GSR, PPG, and facial recordings from 35 participants. FIRMED provides event-centered timestamps, emotion labels, and intensity annotations, and its annotation quality is supported by subjective and physiological validation. Benchmark experiments show that FIRMED consistently outperforms whole-trial labeling, yielding an average gain of 3.8 percentage points across eight EEG-based classifiers, with further improvements under multimodal fusion. FIRMED provides a practical benchmark for temporally localized supervision in multimodal affective computing.

Bohan Li, Jiannan Guan, Longxu Dou et al.

The Myers-Briggs Type Indicator (MBTI) is one of the most influential personality theories reflecting individual differences in thinking, feeling, and behaving. MBTI personality detection has garnered considerable research interest and has evolved significantly over the years. However, this task tends to be overly optimistic, as it currently does not align well with the natural distribution of population personality traits. Specifically, (1) the self-reported labels in existing datasets result in incorrect labeling issues, and (2) the hard labels fail to capture the full range of population personality distributions. In this paper, we optimize the task by constructing MBTIBench, the first manually annotated high-quality MBTI personality detection dataset with soft labels, under the guidance of psychologists. As for the first challenge, MBTIBench effectively solves the incorrect labeling issues, which account for 29.58% of the data. As for the second challenge, we estimate soft labels by deriving the polarity tendency of samples. The obtained soft labels confirm that there are more people with non-extreme personality traits. Experimental results not only highlight the polarized predictions and biases in LLMs as key directions for future research, but also confirm that soft labels can provide more benefits to other psychological tasks than hard labels. The code and data are available at https://github.com/Personality-NLP/MbtiBench.

Bohan Li, Yiming Liu, Xueyan Niu et al.

Diffusion models have achieved remarkable success in generating high quality image and video data. More recently, they have also been used for image compression with high perceptual quality. In this paper, we present a novel approach to extreme video compression leveraging the predictive power of diffusion-based generative models at the decoder. The conditional diffusion model takes several neural compressed frames and generates subsequent frames. When the reconstruction quality drops below the desired level, new frames are encoded to restart prediction. The entire video is sequentially encoded to achieve a visually pleasing reconstruction, considering perceptual quality metrics such as the learned perceptual image patch similarity (LPIPS) and the Frechet video distance (FVD), at bit rates as low as 0.02 bits per pixel (bpp). Experimental results demonstrate the effectiveness of the proposed scheme compared to standard codecs such as H.264 and H.265 in the low bpp regime. The results showcase the potential of exploiting the temporal relations in video data using generative models. Code is available at: https://github.com/ElesionKyrie/Extreme-Video-Compression-With-Prediction-Using-Pre-trainded-Diffusion-Models-

Shaocong Xu, Songlin Wei, Qizhe Wei et al.

Transparent objects remain notoriously hard for perception systems: refraction, reflection and transmission break the assumptions behind stereo, ToF and purely discriminative monocular depth, causing holes and temporally unstable estimates. Our key observation is that modern video diffusion models already synthesize convincing transparent phenomena, suggesting they have internalized the optical rules. We build TransPhy3D, a synthetic video corpus of transparent/reflective scenes: 11k sequences rendered with Blender/Cycles. Scenes are assembled from a curated bank of category-rich static assets and shape-rich procedural assets paired with glass/plastic/metal materials. We render RGB + depth + normals with physically based ray tracing and OptiX denoising. Starting from a large video diffusion model, we learn a video-to-video translator for depth (and normals) via lightweight LoRA adapters. During training we concatenate RGB and (noisy) depth latents in the DiT backbone and co-train on TransPhy3D and existing frame-wise synthetic datasets, yielding temporally consistent predictions for arbitrary-length input videos. The resulting model, DKT, achieves zero-shot SOTA on real and synthetic video benchmarks involving transparency: ClearPose, DREDS (CatKnown/CatNovel), and TransPhy3D-Test. It improves accuracy and temporal consistency over strong image/video baselines, and a normal variant sets the best video normal estimation results on ClearPose. A compact 1.3B version runs at ~0.17 s/frame. Integrated into a grasping stack, DKT's depth boosts success rates across translucent, reflective and diffuse surfaces, outperforming prior estimators. Together, these results support a broader claim: "Diffusion knows transparency." Generative video priors can be repurposed, efficiently and label-free, into robust, temporally coherent perception for challenging real-world manipulation.

Bohan Li, Shengmin Li, Xinyu Shi et al.

Graph Convolutional Networks (GCNs) are widely adopted for tasks involving relational or graph-structured data and can be formulated as two-stage sparse-dense matrix multiplication (SpMM) during inference. However, existing accelerators often struggle with the irregular workloads induced by power-law node degree distributions. In this work, we propose FlexVector, a vector-processor-based architecture that efficiently accelerates SpMM for GCN inference. To address irregular computation patterns, FlexVector adopts a row-wise, product-based dataflow that regularizes SpMM execution and exposes vector parallelism through full-row access to vector registers, eliminating the need for multi-banked register file designs. Building on this dataflow, it introduces software-managed, flexible vector register files (VRFs) that adapt to irregular data access patterns, without sacrificing memory access efficiency. To further exploit these architectural capabilities, we develop a graph-aware preprocessing and node partitioning strategy that restructures irregular graph workloads to better match the row-wise dataflow and VRF capacity. This hardware-software co-design reduces memory traffic, leading to significant performance and energy efficiency gains on real-world GCN workloads. Experimental results on five real-world GCN datasets show that the VRF-centric FlexVector achieves a 3.78x speedup and 40.5% lower energy at comparable area cost relative to a state-of-the-art cache-centric baseline with buffers of the same size.

Ziyang Yan, Ruikai Li, Zhiyong Cui et al.

Online HD map construction is a fundamental task in autonomous driving systems, aiming to acquire semantic information of map elements around the ego vehicle based on real-time sensor inputs. Recently, several approaches have achieved promising results by incorporating offline priors such as SD maps and HD maps or by fusing multi-modal data. However, these methods depend on stale offline maps and multi-modal sensor suites, resulting in avoidable computational overhead at inference. To address these limitations, we employ a knowledge distillation strategy to transfer knowledge from multimodal models with prior knowledge to an efficient, low-cost, and vision-centric student model. Specifically, we propose MapKD, a novel multi-level cross-modal knowledge distillation framework with an innovative Teacher-Coach-Student (TCS) paradigm. This framework consists of: (1) a camera-LiDAR fusion model with SD/HD map priors serving as the teacher; (2) a vision-centric coach model with prior knowledge and simulated LiDAR to bridge the cross-modal knowledge transfer gap; and (3) a lightweight vision-based student model. Additionally, we introduce two targeted knowledge distillation strategies: Token-Guided 2D Patch Distillation (TGPD) for bird's eye view feature alignment and Masked Semantic Response Distillation (MSRD) for semantic learning guidance. Extensive experiments on the challenging nuScenes dataset demonstrate that MapKD improves the student model by +6.68 mIoU and +10.94 mAP while simultaneously accelerating inference speed. The code is available at:https://github.com/2004yan/MapKD2026.

Zixiong Yu, Jun Rao, Guhan Chen et al.

Synthesizing high-quality mathematical reasoning data without human priors remains a significant challenge. Current approaches typically rely on seed data mutation or simple prompt engineering, often suffering from mode collapse and limited logical complexity. This paper proposes a hierarchical synthesis framework that formulates data synthesis as an unsupervised optimization problem over a constraint graph followed by semantic instantiation, rather than treating it as a direct text generation task. We introduce a Legislator-Executor paradigm: The Legislator adversarially evolves structured generation blueprints encoding the constraints of the problem, while the Executor instantiates these specifications into diverse natural language scenarios. This decoupling of skeleton design from linguistic realization enables a prioritized focus on constructing complex and diverse logical structures, thereby guiding high-quality data synthesis. Experiments conducted on a total of 10 models across the Qwen, Llama, Mistral, and Gemma series demonstrate that our method achieves notable results: models fine-tuned on 1K synthesized samples outperform widely-used datasets of comparable scale (LIMO, s1K) across eight mathematical benchmarks, exhibiting superior out-of-distribution generalization.

Xueqing Yu, Bohan Li, Yan Li et al.

Recent Vision-Language Models (VLMs) have made remarkable progress in multimodal understanding tasks, yet their evaluation on long video understanding remains unreliable. Due to limited frame inputs, key frames necessary for answering the question may be missing from the model's input. However, models that truthfully refuse to answer under such uncertainty are marked as incorrect, while those that guess may coincidentally produce the correct answer and thus obtain deceptively higher accuracy, leading to misleading evaluation results and encouraging models to guess rather than respond honestly. To address this issue, we introduce VirtueBench, a benchmark explicitly designed to assess model trustworthiness under uncertainty. VirtueBench constructs multiple frame-sampling levels for each video and provides ground truths that distinguish between answerable and unanswerable cases. Evaluations on 25 open-source and commercial VLMs reveal distinct refusal behaviors across different model families, with refusal accuracy ranging from over 70% in the best models to nearly 0% in the worst. Moreover, most models exhibit a substantial drop in refusal when the prompt does not explicitly require them to do so. These findings highlight the need for developing trustworthy VLMs for multimodal understanding, guided by benchmarks and leaderboards that emphasize reliability and trustworthiness.

Bohan Li, Wenbin Huang, Yuhang Qiu et al.

Large Audio Language Models (LALMs), which couple acoustic perception with large language models (LLMs) to extract and understand diverse information from audio, have attracted intense interest from both academic and industrial communities. However, existing LALMs are highly sensitive to how instructions are phrased, affecting both (i) instruction-following rates and (ii) task performance. Yet, no existing benchmarks offer a systematic and comprehensive evaluation of this sensitivity. We introduce ISA-Bench, a dynamic benchmark evaluating instruction sensitivity for LALMs along three axes: instruction description, output format, and task composition. We assess recent open-source and proprietary LALMs using ISA-Bench, profiling both compliance and accuracy under controlled instruction variations. Experimental results reveal that even state-of-the-art LALMs suffer significant instruction sensitivity, leading to degraded performance on fundamental audio understanding tasks. To mitigate this issue, we fine-tune Qwen2-Audio on a specifically constructed complex instruction-variant dataset, achieving a marked improvement in instruction-following performance. However, this also induces nontrivial catastrophic forgetting: the model loses some previously mastered task capabilities when exposed to new instruction styles. Our benchmark provides a standardized basis for assessing and improving instruction sensitivity in LALMs, underscoring the need for instruction-robust audio understanding in real-world pipelines.

Bohan Li, Xin Jin, Hu Zhu et al.

Driving scene generation is a critical domain for autonomous driving, enabling downstream applications, including perception and planning evaluation. Occupancy-centric methods have recently achieved state-of-the-art results by offering consistent conditioning across frames and modalities; however, their performance heavily depends on annotated occupancy data, which still remains scarce. To overcome this limitation, we curate Nuplan-Occ, the largest semantic occupancy dataset to date, constructed from the widely used Nuplan benchmark. Its scale and diversity facilitate not only large-scale generative modeling but also autonomous driving downstream applications. Based on this dataset, we develop a unified framework that jointly synthesizes high-quality semantic occupancy, multi-view videos, and LiDAR point clouds. Our approach incorporates a spatio-temporal disentangled architecture to support high-fidelity spatial expansion and temporal forecasting of 4D dynamic occupancy. To bridge modal gaps, we further propose two novel techniques: a Gaussian splatting-based sparse point map rendering strategy that enhances multi-view video generation, and a sensor-aware embedding strategy that explicitly models LiDAR sensor properties for realistic multi-LiDAR simulation. Extensive experiments demonstrate that our method achieves superior generation fidelity and scalability compared to existing approaches, and validates its practical value in downstream tasks. Repo: https://github.com/Arlo0o/UniScene-Unified-Occupancy-centric-Driving-Scene-Generation/tree/v2

Wenyao Zhang, Hongsi Liu, Bohan Li et al.

Current self-supervised monocular depth estimation (MDE) approaches encounter performance limitations due to insufficient semantic-spatial knowledge extraction. To address this challenge, we propose Hybrid-depth, a novel framework that systematically integrates foundation models (e.g., CLIP and DINO) to extract visual priors and acquire sufficient contextual information for MDE. Our approach introduces a coarse-to-fine progressive learning framework: 1) Firstly, we aggregate multi-grained features from CLIP (global semantics) and DINO (local spatial details) under contrastive language guidance. A proxy task comparing close-distant image patches is designed to enforce depth-aware feature alignment using text prompts; 2) Next, building on the coarse features, we integrate camera pose information and pixel-wise language alignment to refine depth predictions. This module seamlessly integrates with existing self-supervised MDE pipelines (e.g., Monodepth2, ManyDepth) as a plug-and-play depth encoder, enhancing continuous depth estimation. By aggregating CLIP's semantic context and DINO's spatial details through language guidance, our method effectively addresses feature granularity mismatches. Extensive experiments on the KITTI benchmark demonstrate that our method significantly outperforms SOTA methods across all metrics, which also indeed benefits downstream tasks like BEV perception. Code is available at https://github.com/Zhangwenyao1/Hybrid-depth.

Wenlong Wu, Haofen Wang, Bohan Li et al.

Retrieval Augmented Generation (RAG) has emerged as a promising solution to address hallucination issues in Large Language Models (LLMs). However, the integration of multiple retrieval sources, while potentially more informative, introduces new challenges that can paradoxically exacerbate hallucination problems. These challenges manifest primarily in two aspects: the sparse distribution of multi-source data that hinders the capture of logical relationships and the inherent inconsistencies among different sources that lead to information conflicts. To address these challenges, we propose MultiRAG, a novel framework designed to mitigate hallucination in multi-source retrieval-augmented generation through knowledge-guided approaches. Our framework introduces two key innovations: (1) a knowledge construction module that employs multi-source line graphs to efficiently aggregate logical relationships across different knowledge sources, effectively addressing the sparse data distribution issue; and (2) a sophisticated retrieval module that implements a multi-level confidence calculation mechanism, performing both graph-level and node-level assessments to identify and eliminate unreliable information nodes, thereby reducing hallucinations caused by inter-source inconsistencies. Extensive experiments on four multi-domain query datasets and two multi-hop QA datasets demonstrate that MultiRAG significantly enhances the reliability and efficiency of knowledge retrieval in complex multi-source scenarios. \textcolor{blue}{Our code is available in https://github.com/wuwenlong123/MultiRAG.

Zheng Zhu, Xiaofeng Wang, Wangbo Zhao et al.

General world models represent a crucial pathway toward achieving Artificial General Intelligence (AGI), serving as the cornerstone for various applications ranging from virtual environments to decision-making systems. Recently, the emergence of the Sora model has attained significant attention due to its remarkable simulation capabilities, which exhibits an incipient comprehension of physical laws. In this survey, we embark on a comprehensive exploration of the latest advancements in world models. Our analysis navigates through the forefront of generative methodologies in video generation, where world models stand as pivotal constructs facilitating the synthesis of highly realistic visual content. Additionally, we scrutinize the burgeoning field of autonomous-driving world models, meticulously delineating their indispensable role in reshaping transportation and urban mobility. Furthermore, we delve into the intricacies inherent in world models deployed within autonomous agents, shedding light on their profound significance in enabling intelligent interactions within dynamic environmental contexts. At last, we examine challenges and limitations of world models, and discuss their potential future directions. We hope this survey can serve as a foundational reference for the research community and inspire continued innovation. This survey will be regularly updated at: https://github.com/GigaAI-research/General-World-Models-Survey.

Bohan Li, Hao Zhou, Junxian He et al.

Pre-trained contextual representations like BERT have achieved great success in natural language processing. However, the sentence embeddings from the pre-trained language models without fine-tuning have been found to poorly capture semantic meaning of sentences. In this paper, we argue that the semantic information in the BERT embeddings is not fully exploited. We first reveal the theoretical connection between the masked language model pre-training objective and the semantic similarity task theoretically, and then analyze the BERT sentence embeddings empirically. We find that BERT always induces a non-smooth anisotropic semantic space of sentences, which harms its performance of semantic similarity. To address this issue, we propose to transform the anisotropic sentence embedding distribution to a smooth and isotropic Gaussian distribution through normalizing flows that are learned with an unsupervised objective. Experimental results show that our proposed BERT-flow method obtains significant performance gains over the state-of-the-art sentence embeddings on a variety of semantic textual similarity tasks. The code is available at https://github.com/bohanli/BERT-flow.

Bohan Li, Shuojue Yang, Baorui Peng et al.

Action-conditioned surgical video generation is a critical yet highly challenging problem for robotic surgery. The core difficulty is that low-dimensional control vectors must precisely govern complex image-space evolution. In this work, we propose a kinematic-to-visual lifting paradigm that converts articulated kinematics into a unified set of five image-aligned control modalities. Building on this representation, we introduce a hierarchically routed visual control framework that selectively activates the most relevant control modalities and motion scales. Instead of uniformly applying all control signals, our model performs hierarchical routing to dynamically allocate conditioning capacity. We further design kinematic-prior-guided routing loss functions to ensure physically meaningful, temporally stable, and efficient expert utilization. To improve efficiency, we propose a budgeted training and inference scheme that leverages routing-induced sparsity. By selectively discarding low-significance control pathways during training and execution, our approach enables adaptive computation that is complementary to standard distillation. We additionally construct a new benchmark with curated articulated annotations, obtained through human-in-the-loop semantic labeling and differentiable pose tracking, providing realistic supervision for action-conditioned surgical video generation. Extensive experiments demonstrate that our method consistently improves action faithfulness, visual fidelity, and cross-domain generalization over diverse baselines. Moreover, our efficient variant achieves substantial reductions in latency while maintaining strong control accuracy.

Mingju Gao, Kaisen Yang, Huan-ang Gao et al.

Hand-object interaction (HOI) reconstruction and synthesis are becoming central to embodied AI and AR/VR. Yet, despite rapid progress, existing HOI generation research remains fragmented across three disjoint tracks: (1) pose-only synthesis that predicts MANO trajectories without producing pixels; (2) single-image HOI generation that hallucinates appearance from masks or 2D cues but lacks dynamics; and (3) video generation methods that require both the entire pose sequence and the ground-truth first frame as inputs, preventing true sim-to-real deployment. Inspired by the philosophy of Joo et al. (2018), we think that HOI generation requires a unified engine that brings together pose, appearance, and motion within one coherent framework. Thus we introduce PAM: a Pose-Appearance-Motion Engine for controllable HOI video generation. The performance of our engine is validated by: (1) On DexYCB, we obtain an FVD of 29.13 (vs. 38.83 for InterDyn), and MPJPE of 19.37 mm (vs. 30.05 mm for CosHand), while generating higher-resolution 480x720 videos compared to 256x256 and 256x384 baselines. (2) On OAKINK2, our full multi-condition model improves FVD from 68.76 to 46.31. (3) An ablation over input conditions on DexYCB shows that combining depth, segmentation, and keypoints consistently yields the best results. (4) For a downstream hand pose estimation task using SimpleHand, augmenting training with 3,400 synthetic videos (207k frames) allows a model trained on only 50% of the real data plus our synthetic data to match the 100% real baseline.

Bohan Li, Jiazhe Guo, Hongsi Liu et al.

Generating high-fidelity, controllable, and annotated training data is critical for autonomous driving. Existing methods typically generate a single data form directly from a coarse scene layout, which not only fails to output rich data forms required for diverse downstream tasks but also struggles to model the direct layout-to-data distribution. In this paper, we introduce UniScene, the first unified framework for generating three key data forms - semantic occupancy, video, and LiDAR - in driving scenes. UniScene employs a progressive generation process that decomposes the complex task of scene generation into two hierarchical steps: (a) first generating semantic occupancy from a customized scene layout as a meta scene representation rich in both semantic and geometric information, and then (b) conditioned on occupancy, generating video and LiDAR data, respectively, with two novel transfer strategies of Gaussian-based Joint Rendering and Prior-guided Sparse Modeling. This occupancy-centric approach reduces the generation burden, especially for intricate scenes, while providing detailed intermediate representations for the subsequent generation stages. Extensive experiments demonstrate that UniScene outperforms previous SOTAs in the occupancy, video, and LiDAR generation, which also indeed benefits downstream driving tasks. Project page: https://arlo0o.github.io/uniscene/

Yiwei Guo, Zhihan Li, Hankun Wang et al.

The rapid advancement of speech generation technologies in the era of large language models (LLMs) has established discrete speech tokens as a foundational paradigm for speech representation. These tokens, characterized by their discrete, compact, and concise nature, are not only advantageous for efficient transmission and storage, but also inherently compatible with the language modeling framework, enabling seamless integration of speech into text-dominated LLM architectures. Current research categorizes discrete speech tokens into two principal classes: acoustic tokens and semantic tokens, each of which has evolved into a rich research domain characterized by unique design philosophies and methodological approaches. This survey systematically synthesizes the existing taxonomy and recent innovations in discrete speech tokenization, conducts a critical examination of the strengths and limitations of each paradigm, and presents systematic experimental comparisons across token types. Furthermore, we identify persistent challenges in the field and propose potential research directions, aiming to offer actionable insights to inspire future advancements in the development and application of discrete speech tokens.

Jiazhe Guo, Yikang Ding, Xiwu Chen et al.

Current generative models struggle to synthesize dynamic 4D driving scenes that simultaneously support temporal extrapolation and spatial novel view synthesis (NVS) without per-scene optimization. A key challenge lies in finding an efficient and generalizable geometric representation that seamlessly connects temporal and spatial synthesis. To address this, we propose DiST-4D, the first disentangled spatiotemporal diffusion framework for 4D driving scene generation, which leverages metric depth as the core geometric representation. DiST-4D decomposes the problem into two diffusion processes: DiST-T, which predicts future metric depth and multi-view RGB sequences directly from past observations, and DiST-S, which enables spatial NVS by training only on existing viewpoints while enforcing cycle consistency. This cycle consistency mechanism introduces a forward-backward rendering constraint, reducing the generalization gap between observed and unseen viewpoints. Metric depth is essential for both accurate reliable forecasting and accurate spatial NVS, as it provides a view-consistent geometric representation that generalizes well to unseen perspectives. Experiments demonstrate that DiST-4D achieves state-of-the-art performance in both temporal prediction and NVS tasks, while also delivering competitive performance in planning-related evaluations.

Hao Li, Roy Qin, Zhengyu Zou et al.

Applying Gaussian Splatting to perception tasks for 3D scene understanding is becoming increasingly popular. Most existing works primarily focus on rendering 2D feature maps from novel viewpoints, which leads to an imprecise 3D language field with outlier languages, ultimately failing to align objects in 3D space. By utilizing masked images for feature extraction, these approaches also lack essential contextual information, leading to inaccurate feature representation. To this end, we propose a Language-Embedded Surface Field (LangSurf), which accurately aligns the 3D language fields with the surface of objects, facilitating precise 2D and 3D segmentation with text query, widely expanding the downstream tasks such as removal and editing. The core of LangSurf is a joint training strategy that flattens the language Gaussian on the object surfaces using geometry supervision and contrastive losses to assign accurate language features to the Gaussians of objects. In addition, we also introduce the Hierarchical-Context Awareness Module to extract features at the image level for contextual information then perform hierarchical mask pooling using masks segmented by SAM to obtain fine-grained language features in different hierarchies. Extensive experiments on open-vocabulary 2D and 3D semantic segmentation demonstrate that LangSurf outperforms the previous state-of-the-art method LangSplat by a large margin. As shown in Fig. 1, our method is capable of segmenting objects in 3D space, thus boosting the effectiveness of our approach in instance recognition, removal, and editing, which is also supported by comprehensive experiments. \url{https://langsurf.github.io}.

Wenbin Huang, Yuhang Qiu, Bohan Li et al.

Automatic speech recognition systems often produce confident yet incorrect transcriptions under noisy or ambiguous conditions, which can be misleading for both users and downstream applications. Standard evaluation based on Word Error Rate focuses solely on accuracy and fails to capture transcription reliability. We introduce an abstention-aware transcription framework that enables ASR models to explicitly abstain from uncertain segments. To evaluate reliability under abstention, we propose RAS, a reliability-oriented metric that balances transcription informativeness and error aversion, with its trade-off parameter calibrated by human preference. We then train an abstention-aware ASR model through supervised bootstrapping followed by reinforcement learning. Our experiments demonstrate substantial improvements in transcription reliability while maintaining competitive accuracy.

Bohan Li, Hankun Wang, Situo Zhang et al.

The auto-regressive architecture, like GPTs, is widely used in modern Text-to-Speech (TTS) systems. However, it incurs substantial inference time, particularly due to the challenges in the next-token prediction posed by lengthy sequences of speech tokens. In this work, we introduce VADUSA, one of the first approaches to accelerate auto-regressive TTS through speculative decoding. Our results show that VADUSA not only significantly improves inference speed but also enhances performance by incorporating draft heads to predict future speech content auto-regressively. Furthermore, the inclusion of a tolerance mechanism during sampling accelerates inference without compromising quality. Our approach demonstrates strong generalization across large datasets and various types of speech tokens.

Xin Jin, Bohan Li, BAAO Xie et al.

Representation disentanglement may help AI fundamentally understand the real world and thus benefit both discrimination and generation tasks. It currently has at least three unresolved core issues: (i) heavy reliance on label annotation and synthetic data -- causing poor generalization on natural scenarios; (ii) heuristic/hand-craft disentangling constraints make it hard to adaptively achieve an optimal training trade-off; (iii) lacking reasonable evaluation metric, especially for the real label-free data. To address these challenges, we propose a \textbf{C}losed-\textbf{L}oop unsupervised representation \textbf{Dis}entanglement approach dubbed \textbf{CL-Dis}. Specifically, we use diffusion-based autoencoder (Diff-AE) as a backbone while resorting to $β$-VAE as a co-pilot to extract semantically disentangled representations. The strong generation ability of diffusion model and the good disentanglement ability of VAE model are complementary. To strengthen disentangling, VAE-latent distillation and diffusion-wise feedback are interconnected in a closed-loop system for a further mutual promotion. Then, a self-supervised \textbf{Navigation} strategy is introduced to identify interpretable semantic directions in the disentangled latent space. Finally, a new metric based on content tracking is designed to evaluate the disentanglement effect. Experiments demonstrate the superiority of CL-Dis on applications like real image manipulation and visual analysis.

Zhiyuan Xu, Bohan Li, Huan-ang Gao et al.

Generating photorealistic driving videos has seen significant progress recently, but current methods largely focus on ordinary, non-adversarial scenarios. Meanwhile, efforts to generate adversarial driving scenarios often operate on abstract trajectory or BEV representations, falling short of delivering realistic sensor data that can truly stress-test autonomous driving (AD) systems. In this work, we introduce Challenger, a framework that produces physically plausible yet photorealistic adversarial driving videos. Generating such videos poses a fundamental challenge: it requires jointly optimizing over the space of traffic interactions and high-fidelity sensor observations. Challenger makes this affordable through two techniques: (1) a physics-aware multi-round trajectory refinement process that narrows down candidate adversarial maneuvers, and (2) a tailored trajectory scoring function that encourages realistic yet adversarial behavior while maintaining compatibility with downstream video synthesis. As tested on the nuScenes dataset, Challenger generates a diverse range of aggressive driving scenarios-including cut-ins, sudden lane changes, tailgating, and blind spot intrusions-and renders them into multiview photorealistic videos. Extensive evaluations show that these scenarios significantly increase the collision rate of state-of-the-art end-to-end AD models (UniAD, VAD, SparseDrive, and DiffusionDrive), and importantly, adversarial behaviors discovered for one model often transfer to others.

Nan Wang, Yuantao Chen, Lixing Xiao et al.

Neural rendering techniques, including NeRF and Gaussian Splatting (GS), rely on photometric consistency to produce high-quality reconstructions. However, in real-world scenarios, it is challenging to guarantee perfect photometric consistency in acquired images. Appearance codes have been widely used to address this issue, but their modeling capability is limited, as a single code is applied to the entire image. Recently, the bilateral grid was introduced to perform pixel-wise color mapping, but it is difficult to optimize and constrain effectively. In this paper, we propose a novel multi-scale bilateral grid that unifies appearance codes and bilateral grids. We demonstrate that this approach significantly improves geometric accuracy in dynamic, decoupled autonomous driving scene reconstruction, outperforming both appearance codes and bilateral grids. This is crucial for autonomous driving, where accurate geometry is important for obstacle avoidance and control. Our method shows strong results across four datasets: Waymo, NuScenes, Argoverse, and PandaSet. We further demonstrate that the improvement in geometry is driven by the multi-scale bilateral grid, which effectively reduces floaters caused by photometric inconsistency.

Xiuyu Yang, Bohan Li, Shaocong Xu et al.

Acquiring real-world robotic simulation data through teleoperation is notoriously time-consuming and labor-intensive. Recently, action-driven generative models have gained widespread adoption in robot learning and simulation, as they eliminate safety concerns and reduce maintenance efforts. However, the action sequences used in these methods often result in limited control precision and poor generalization due to their globally coarse alignment. To address these limitations, we propose ORV, an Occupancy-centric Robot Video generation framework, which utilizes 4D semantic occupancy sequences as a fine-grained representation to provide more accurate semantic and geometric guidance for video generation. By leveraging occupancy-based representations, ORV enables seamless translation of simulation data into photorealistic robot videos, while ensuring high temporal consistency and precise controllability. Furthermore, our framework supports the simultaneous generation of multi-view videos of robot gripping operations - an important capability for downstream robotic learning tasks. Extensive experimental results demonstrate that ORV consistently outperforms existing baseline methods across various datasets and sub-tasks. Demo, Code and Model: https://orangesodahub.github.io/ORV

Zheng Geng, Nan Wang, Shaocong Xu et al.

Estimating the 6D pose of arbitrary unseen objects from a single reference image is critical for robotics operating in the long-tail of real-world instances. However, this setting is notoriously challenging: 3D models are rarely available, single-view reconstructions lack metric scale, and domain gaps between generated models and real-world images undermine robustness. We propose OnePoseViaGen, a pipeline that tackles these challenges through two key components. First, a coarse-to-fine alignment module jointly refines scale and pose by combining multi-view feature matching with render-and-compare refinement. Second, a text-guided generative domain randomization strategy diversifies textures, enabling effective fine-tuning of pose estimators with synthetic data. Together, these steps allow high-fidelity single-view 3D generation to support reliable one-shot 6D pose estimation. On challenging benchmarks (YCBInEOAT, Toyota-Light, LM-O), OnePoseViaGen achieves state-of-the-art performance far surpassing prior approaches. We further demonstrate robust dexterous grasping with a real robot hand, validating the practicality of our method in real-world manipulation. Project page: https://gzwsama.github.io/OnePoseviaGen.github.io/