arXiv subjects

Xin Wang

Publications and source records attributed to Xin Wang.

At least 721 records · Page 40Linked to original sources

arXiv2025

Embodied-R: Collaborative Framework for Activating Embodied Spatial Reasoning in Foundation Models via Reinforcement Learning

Humans can perceive and reason about spatial relationships from sequential visual observations, such as egocentric video streams. However, how pretrained models acquire such abilities, especially high-level reasoning, remains unclear. This paper introduces Embodied-R, a collaborative framework combining large-scale Vision-Language Models (VLMs) for perception and small-scale Language Models (LMs) for reasoning. Using Reinforcement Learning (RL) with a novel reward system considering think-answer logical consistency, the model achieves slow-thinking capabilities with limited computational resources. After training on only 5k embodied video samples, Embodied-R with a 3B LM matches state-of-the-art multimodal reasoning models (OpenAI-o1, Gemini-2.5-pro) on both in-distribution and out-of-distribution embodied spatial reasoning tasks. Embodied-R also exhibits emergent thinking patterns such as systematic analysis and contextual integration. We further explore research questions including response length, training on VLM, strategies for reward design, and differences in model generalization after SFT (Supervised Fine-Tuning) and RL training.

cs.AI↗

arXiv2025

Diffusion-empowered AutoPrompt MedSAM

MedSAM, a medical foundation model derived from the SAM architecture, has demonstrated notable success across diverse medical domains. However, its clinical application faces two major challenges: the dependency on labor-intensive manual prompt generation, which imposes a significant burden on clinicians, and the absence of semantic labeling in the generated segmentation masks for organs or lesions, limiting its practicality for non-expert users. To address these limitations, we propose AutoMedSAM, an end-to-end framework derived from SAM, designed to enhance usability and segmentation performance. AutoMedSAM retains MedSAM's image encoder and mask decoder structure while introducing a novel diffusion-based class prompt encoder. The diffusion-based encoder employs a dual-decoder structure to collaboratively generate prompt embeddings guided by sparse and dense prompt definitions. These embeddings enhance the model's ability to understand and process clinical imagery autonomously. With this encoder, AutoMedSAM leverages class prompts to embed semantic information into the model's predictions, transforming MedSAM's semi-automated pipeline into a fully automated workflow. Furthermore, AutoMedSAM employs an uncertainty-aware joint optimization strategy during training to effectively inherit MedSAM's pre-trained knowledge while improving generalization by integrating multiple loss functions. Experimental results across diverse datasets demonstrate that AutoMedSAM achieves superior performance while broadening its applicability to both clinical settings and non-expert users. Code is available at https://github.com/HP-ML/AutoPromptMedSAM.git.

eess.IV↗

arXiv2025

Modular-Cam: Modular Dynamic Camera-view Video Generation with LLM

Text-to-Video generation, which utilizes the provided text prompt to generate high-quality videos, has drawn increasing attention and achieved great success due to the development of diffusion models recently. Existing methods mainly rely on a pre-trained text encoder to capture the semantic information and perform cross attention with the encoded text prompt to guide the generation of video. However, when it comes to complex prompts that contain dynamic scenes and multiple camera-view transformations, these methods can not decompose the overall information into separate scenes, as well as fail to smoothly change scenes based on the corresponding camera-views. To solve these problems, we propose a novel method, i.e., Modular-Cam. Specifically, to better understand a given complex prompt, we utilize a large language model to analyze user instructions and decouple them into multiple scenes together with transition actions. To generate a video containing dynamic scenes that match the given camera-views, we incorporate the widely-used temporal transformer into the diffusion model to ensure continuity within a single scene and propose CamOperator, a modular network based module that well controls the camera movements. Moreover, we propose AdaControlNet, which utilizes ControlNet to ensure consistency across scenes and adaptively adjusts the color tone of the generated video. Extensive qualitative and quantitative experiments prove our proposed Modular-Cam's strong capability of generating multi-scene videos together with its ability to achieve fine-grained control of camera movements. Generated results are available at https://modular-cam.github.io.

cs.CV↗

arXiv2025

CAPERS Observations of Two UV-Bright Galaxies at z>10. More Evidence for Bursting Star Formation in the Early Universe

We present the first results from the CAPERS survey, utilizing PRISM observations with the JWST/NIRSpec MSA in the PRIMER-UDS field. With just 14 % of the total planned data volume, we spectroscopically confirm two new bright galaxies ($M_{\rm UV}\sim -20.4$) at redshifts $z = 10.562\pm0.034$ and $z = 11.013\pm0.028$. We examine their physical properties, morphologies, and star formation histories, finding evidence for recent bursting star formation in at least one galaxy thanks to the detection of strong (EW$_0\sim70$ A) H$γ$ emission. Combining our findings with previous studies of similarly bright objects at high-$z$, we further assess the role of stochastic star formation processes in shaping early galaxy populations. Our analysis finds that the majority of bright ($M_{\rm UV}\lesssim -20$) spectroscopically-confirmed galaxies at $z>10$ were likely observed during a starburst episode, characterized by a median SFR$_{10}$/SFR$_{100}\sim2$, although with substantial scatter. Our work also finds tentative evidence that $z>10$ galaxies are more preferentially in a bursting phase than similarly bright $z\sim6$ galaxies. We finally discuss the prospects of deeper spectroscopic observations of a statistically significant number of bright galaxies to quantify the true impact of bursting star formation on the evolution of the bright end of the ultraviolet luminosity function at these early epochs.

astro-ph.GA↗

arXiv2025

SpoofCeleb: Speech Deepfake Detection and SASV In The Wild

This paper introduces SpoofCeleb, a dataset designed for Speech Deepfake Detection (SDD) and Spoofing-robust Automatic Speaker Verification (SASV), utilizing source data from real-world conditions and spoofing attacks generated by Text-To-Speech (TTS) systems also trained on the same real-world data. Robust recognition systems require speech data recorded in varied acoustic environments with different levels of noise to be trained. However, current datasets typically include clean, high-quality recordings (bona fide data) due to the requirements for TTS training; studio-quality or well-recorded read speech is typically necessary to train TTS models. Current SDD datasets also have limited usefulness for training SASV models due to insufficient speaker diversity. SpoofCeleb leverages a fully automated pipeline we developed that processes the VoxCeleb1 dataset, transforming it into a suitable form for TTS training. We subsequently train 23 contemporary TTS systems. SpoofCeleb comprises over 2.5 million utterances from 1,251 unique speakers, collected under natural, real-world conditions. The dataset includes carefully partitioned training, validation, and evaluation sets with well-controlled experimental protocols. We present the baseline results for both SDD and SASV tasks. All data, protocols, and baselines are publicly available at https://jungjee.github.io/spoofceleb.

cs.SD↗

arXiv2025

VideoDreamer: Customized Multi-Subject Text-to-Video Generation with Disen-Mix Finetuning on Language-Video Foundation Models

Customized text-to-video generation aims to generate text-guided videos with user-given subjects, which has gained increasing attention. However, existing works are primarily limited to single-subject oriented text-to-video generation, leaving the more challenging problem of customized multi-subject generation unexplored. In this paper, we fill this gap and propose a novel VideoDreamer framework, which can generate temporally consistent text-guided videos that faithfully preserve the visual features of the given multiple subjects. Specifically, VideoDreamer adopts the pretrained Stable Diffusion with temporal modules as its base video generator, taking the power of the text-to-image model to generate diversified content. The video generator is further customized for multi-subjects, which leverages the proposed Disen-Mix Finetuning and Human-in-the-Loop Re-finetuning strategy, to tackle the attribute binding problem of multi-subject generation. Additionally, we present a disentangled motion customization strategy to finetune the temporal modules so that we can generate videos with both customized subjects and motions. To evaluate the performance of customized multi-subject text-to-video generation, we introduce the MultiStudioBench benchmark. Extensive experiments demonstrate the remarkable ability of VideoDreamer to generate videos with new content such as new events and backgrounds, tailored to the customized multiple subjects.

cs.CV↗

arXiv2025

Complexity-Scalable Near-Optimal Transceiver Design for Massive MIMO-BICM Systems

Future wireless networks are envisioned to employ multiple-input multiple-output (MIMO) transmissions with large array sizes, and therefore, the adoption of complexity-scalable transceiver becomes important. In this paper, we propose a novel complexity-scalable transceiver design for MIMO systems exploiting bit-interleaved coded modulation (termed MIMO-BICM systems). The proposed scheme leverages the channel bidiagonalization decomposition (CBD), based on which an optimization framework for the precoder and post-processor is developed for maximizing the mutual information (MI) with finite-alphabet inputs. Particularly, we unveil that the desired precoder and post-processor behave distinctively with respect to the operating signal-to-noise ratio (SNR), where the equivalent channel condition number (ECCN) serves as an effective indicator for the overall achievable rate performance. Specifically, at low SNRs, diagonal transmission with a large ECCN is advantageous, while at high SNRs, uniform subchannel gains with a small ECCN are preferred. This allows us to further propose a low-complexity generalized parallel CBD design (GP-CBD) based on Givens rotation according to a well-approximated closed-form performance metric on the achievable rates that takes into account the insights from the ECCN. Numerical results validate the superior performance of the proposed scheme in terms of achievable rate and bit error rate (BER), compared to state-of-the-art designs across various modulation and coding schemes (MCSs).

eess.SP↗

arXiv2025

FAST Drift Scan Survey for HI Intensity Mapping. II. Stacking-based Beam Construction of the 19-feed Array at $1.4$ GHz

Neutral hydrogen (HI) intensity mapping (IM) presents great promise for future cosmological large-scale structure surveys. However, a major challenge for HIIM cosmological studies is to accurately subtract the foreground contamination. An accurate beam model is crucial for improving the quality of foreground subtraction. In this work, we develop a stacking-based beam reconstruction method utilizing the radio continuum point sources within the drift-scan field. Based on the Five-hundred-meter Aperture Spherical radio Telescope (FAST), we employ two sets of drift-scan survey data and merge the measurements to construct the beam patterns of the 19 FAST L-band feeds. To model the beams, we utilize the Zernike polynomial (ZP), which effectively captures asymmetric features of the main beam and the different side lobes. Due to the symmetric location of the beams, the main features of the beams are closely related to the distance from the center of the feed array, e.g., as the distance increases, side lobes become more pronounced. This modeling pipeline leverages the stable drift-scan data to extract beam patterns while accounting for and excluding the reflector's changing effects. It provides a more accurate measurement beam and a more precise model beam for FAST HIIM cosmology surveys.

astro-ph.IM↗

arXiv2025

Recent star formation in 0.5<z<1.5 quiescent galaxies

Observations of massive, quiescent galaxies reveal a relatively uniform evolution: following prolific star formation in the early universe, these galaxies quench and transition to their characteristic quiescent state in the local universe. The debate on the relative role and frequency of the process(es) driving this evolution is robust. In this letter, we identify 0.5<z<1.5 massive, quiescent galaxies in the HST/UVCANDELS extragalactic deep fields using traditional color selection methods and model their spectral energy distributions, which incorporates novel UV images. This analysis reveals ~15% of massive, quiescent galaxies have experienced minor, recent star formation(<10% of total stellar mass within the past ~1Gyr). We find only a marginal, positive correlation between the probability for recent star formation and a measure of the richness of the local environment from a statistical analysis. Assuming the recent star formation present in these quiescent galaxies is physically linked to the local environment, these results suggest only a minor role for dynamic external processes (galaxy mergers and interactions) in the formation and evolution of these galaxies at this redshift.

astro-ph.GA↗

arXiv2025

Mixed-Precision Quantization for Deep Vision Models with Integer Quadratic Programming

Quantization is a widely used technique to compress neural networks. Assigning uniform bit-widths across all layers can result in significant accuracy degradation at low precision and inefficiency at high precision. Mixed-precision quantization (MPQ) addresses this by assigning varied bit-widths to layers, optimizing the accuracy-efficiency trade-off. Existing sensitivity-based methods for MPQ assume that quantization errors across layers are independent, which leads to suboptimal choices. We introduce CLADO, a practical sensitivity-based MPQ algorithm that captures cross-layer dependency of quantization error. CLADO approximates pairwise cross-layer errors using linear equations on a small data subset. Layerwise bit-widths are assigned by optimizing a new MPQ formulation based on cross-layer quantization errors using an Integer Quadratic Program. Experiments with CNN and vision transformer models on ImageNet demonstrate that CLADO achieves state-of-the-art mixed-precision quantization performance. Code repository available here: https://github.com/JamesTuna/CLADO_MPQ

cs.NE↗

arXiv2025

Large Language Model Enhanced Knowledge Representation Learning: A Survey

Knowledge Representation Learning (KRL) is crucial for enabling applications of symbolic knowledge from Knowledge Graphs (KGs) to downstream tasks by projecting knowledge facts into vector spaces. Despite their effectiveness in modeling KG structural information, KRL methods are suffering from the sparseness of KGs. The rise of Large Language Models (LLMs) built on the Transformer architecture presents promising opportunities for enhancing KRL by incorporating textual information to address information sparsity in KGs. LLM-enhanced KRL methods, including three key approaches, encoder-based methods that leverage detailed contextual information, encoder-decoder-based methods that utilize a unified Seq2Seq model for comprehensive encoding and decoding, and decoder-based methods that utilize extensive knowledge from large corpora, have significantly advanced the effectiveness and generalization of KRL in addressing a wide range of downstream tasks. This work provides a broad overview of downstream tasks while simultaneously identifying emerging research directions in these evolving domains.

cs.CL↗

arXiv2025

Unifying Unsupervised Graph-Level Anomaly Detection and Out-of-Distribution Detection: A Benchmark

To build safe and reliable graph machine learning systems, unsupervised graph-level anomaly detection (GLAD) and unsupervised graph-level out-of-distribution (OOD) detection (GLOD) have received significant attention in recent years. Though those two lines of research indeed share the same objective, they have been studied independently in the community due to distinct evaluation setups, creating a gap that hinders the application and evaluation of methods from one to the other. To bridge the gap, in this work, we present a \underline{\textbf{U}}nified \underline{\textbf{B}}enchmark for unsupervised \underline{\textbf{G}}raph-level \underline{\textbf{O}}OD and anoma\underline{\textbf{L}}y \underline{\textbf{D}}etection (\ourmethod), a comprehensive evaluation framework that unifies GLAD and GLOD under the concept of generalized graph-level OOD detection. Our benchmark encompasses 35 datasets spanning four practical anomaly and OOD detection scenarios, facilitating the comparison of 18 representative GLAD/GLOD methods. We conduct multi-dimensional analyses to explore the effectiveness, OOD sensitivity spectrum, robustness, and efficiency of existing methods, shedding light on their strengths and limitations. Furthermore, we provide an open-source codebase (https://github.com/UB-GOLD/UB-GOLD) of \ourmethod to foster reproducible research and outline potential directions for future investigations based on our insights.

cs.LG↗

arXiv2025

Trustworthy Enhanced Multi-view Multi-modal Alzheimer's Disease Prediction with Brain-wide Imaging Transcriptomics Data

Brain transcriptomics provides insights into the molecular mechanisms by which the brain coordinates its functions and processes. However, existing multimodal methods for predicting Alzheimer's disease (AD) primarily rely on imaging and sometimes genetic data, often neglecting the transcriptomic basis of brain. Furthermore, while striving to integrate complementary information between modalities, most studies overlook the informativeness disparities between modalities. Here, we propose TMM, a trusted multiview multimodal graph attention framework for AD diagnosis, using extensive brain-wide transcriptomics and imaging data. First, we construct view-specific brain regional co-function networks (RRIs) from transcriptomics and multimodal radiomics data to incorporate interaction information from both biomolecular and imaging perspectives. Next, we apply graph attention (GAT) processing to each RRI network to produce graph embeddings and employ cross-modal attention to fuse transcriptomics-derived embedding with each imagingderived embedding. Finally, a novel true-false-harmonized class probability (TFCP) strategy is designed to assess and adaptively adjust the prediction confidence of each modality for AD diagnosis. We evaluate TMM using the AHBA database with brain-wide transcriptomics data and the ADNI database with three imaging modalities (AV45-PET, FDG-PET, and VBM-MRI). The results demonstrate the superiority of our method in identifying AD, EMCI, and LMCI compared to state-of-the-arts. Code and data are available at https://github.com/Yaolab-fantastic/TMM.

cs.AI↗

arXiv2025

Dual-stream Transformer-GCN Model with Contextualized Representations Learning for Monocular 3D Human Pose Estimation

This paper introduces a novel approach to monocular 3D human pose estimation using contextualized representation learning with the Transformer-GCN dual-stream model. Monocular 3D human pose estimation is challenged by depth ambiguity, limited 3D-labeled training data, imbalanced modeling, and restricted model generalization. To address these limitations, our work introduces a groundbreaking motion pre-training method based on contextualized representation learning. Specifically, our method involves masking 2D pose features and utilizing a Transformer-GCN dual-stream model to learn high-dimensional representations through a self-distillation setup. By focusing on contextualized representation learning and spatial-temporal modeling, our approach enhances the model's ability to understand spatial-temporal relationships between postures, resulting in superior generalization. Furthermore, leveraging the Transformer-GCN dual-stream model, our approach effectively balances global and local interactions in video pose estimation. The model adaptively integrates information from both the Transformer and GCN streams, where the GCN stream effectively learns local relationships between adjacent key points and frames, while the Transformer stream captures comprehensive global spatial and temporal features. Our model achieves state-of-the-art performance on two benchmark datasets, with an MPJPE of 38.0mm and P-MPJPE of 31.9mm on Human3.6M, and an MPJPE of 15.9mm on MPI-INF-3DHP. Furthermore, visual experiments on public datasets and in-the-wild videos demonstrate the robustness and generalization capabilities of our approach.

cs.CV↗

arXiv2025

Hypothesis testing of symmetry in quantum dynamics

Symmetry plays a crucial role in quantum physics, dictating the behavior and dynamics of physical systems. In this paper, we develop a hypothesis-testing framework for quantum dynamics symmetry using a limited number of queries to the unknown unitary operation and establish the quantum max-relative entropy lower bound for the type-II error. We construct optimal ancilla-free protocols that achieve optimal type-II error probability for testing time-reversal symmetry (T-symmetry) and diagonal symmetry (Z-symmetry) with limited queries. Contrasting with the advantages of indefinite causal order strategies in various quantum information processing tasks, we show that parallel, adaptive, and indefinite causal order strategies have equal power for our tasks. We establish optimal protocols for T-symmetry testing and Z-symmetry testing for 6 and 5 queries, respectively, from which we infer that the type-II error exhibits a decay rate of $\mathcal{O}(m^{-2})$ with respect to the number of queries $m$. This represents a significant improvement over the basic repetition protocols without using global entanglement, where the error decays at a slower rate of $\mathcal{O}(m^{-1})$.

quant-ph↗

arXiv2025

A Benchmark for Multi-speaker Anonymization

Privacy-preserving voice protection approaches primarily suppress privacy-related information derived from paralinguistic attributes while preserving the linguistic content. Existing solutions focus particularly on single-speaker scenarios. However, they lack practicality for real-world applications, i.e., multi-speaker scenarios. In this paper, we present an initial attempt to provide a multi-speaker anonymization benchmark by defining the task and evaluation protocol, proposing benchmarking solutions, and discussing the privacy leakage of overlapping conversations. The proposed benchmark solutions are based on a cascaded system that integrates spectral-clustering-based speaker diarization and disentanglement-based speaker anonymization using a selection-based anonymizer. To improve utility, the benchmark solutions are further enhanced by two conversation-level speaker vector anonymization methods. The first method minimizes the differential similarity across speaker pairs in the original and anonymized conversations, which maintains original speaker relationships in the anonymized version. The other minimizes the aggregated similarity across anonymized speakers, which achieves better differentiation between speakers.Experiments conducted on both non-overlap simulated and real-world datasets demonstrate the effectiveness of the multi-speaker anonymization system with the proposed speaker anonymizers. Additionally, we analyzed overlapping speech regarding privacy leakage and provided potential solutions

cs.SD↗

arXiv2025

Towards Better Alignment: Training Diffusion Models with Reinforcement Learning Against Sparse Rewards

Diffusion models have achieved remarkable success in text-to-image generation. However, their practical applications are hindered by the misalignment between generated images and corresponding text prompts. To tackle this issue, reinforcement learning (RL) has been considered for diffusion model fine-tuning. Yet, RL's effectiveness is limited by the challenge of sparse reward, where feedback is only available at the end of the generation process. This makes it difficult to identify which actions during the denoising process contribute positively to the final generated image, potentially leading to ineffective or unnecessary denoising policies. To this end, this paper presents a novel RL-based framework that addresses the sparse reward problem when training diffusion models. Our framework, named $\text{B}^2\text{-DiffuRL}$, employs two strategies: \textbf{B}ackward progressive training and \textbf{B}ranch-based sampling. For one thing, backward progressive training focuses initially on the final timesteps of denoising process and gradually extends the training interval to earlier timesteps, easing the learning difficulty from sparse rewards. For another, we perform branch-based sampling for each training interval. By comparing the samples within the same branch, we can identify how much the policies of the current training interval contribute to the final image, which helps to learn effective policies instead of unnecessary ones. $\text{B}^2\text{-DiffuRL}$ is compatible with existing optimization algorithms. Extensive experiments demonstrate the effectiveness of $\text{B}^2\text{-DiffuRL}$ in improving prompt-image alignment and maintaining diversity in generated images. The code for this work is available.

cs.CV↗

arXiv2025

Neutrino Theory in the Precision Era

This document summarises discussions on future directions in theoretical neutrino physics, which are the outcome of a neutrino theory workshop held at CERN in February 2025. The starting point is the realisation that neutrino physics offers unique opportunities to address some of the most fundamental questions in physics. This motivates a vigorous experimental programme which the theory community fully supports. \textbf{A strong effort in theoretical neutrino physics is paramount to optimally take advantage of upcoming neutrino experiments and to explore the synergies with other areas of particle, astroparticle, and nuclear physics, as well as cosmology.} Progress on the theory side has the potential to significantly boost the physics reach of experiments, as well as go well beyond their original scope. Strong collaboration between theory and experiment is essential in the precision era. To foster such collaboration, \textbf{we propose to establish a CERN Neutrino Physics Centre.} Taking inspiration from the highly successful LHC Physics Center at Fermilab, the CERN Neutrino Physics Centre would be the European hub of the neutrino community, covering experimental and theoretical activities.

hep-ph↗