arXiv subjects

Cheng Yuan

Publications and source records attributed to Cheng Yuan.

At least 37 records · Page 2Linked to original sources

arXiv2026

Privacy-Aware Camera 2.0 Technical Report

With the increasing deployment of intelligent sensing technologies in highly sensitive environments such as restrooms and locker rooms, visual surveillance systems face a profound privacy-security paradox. Existing privacy-preserving approaches, including physical desensitization, encryption, and obfuscation, often compromise semantic understanding or fail to ensure mathematically provable irreversibility. Although Privacy Camera 1.0 eliminated visual data at the source to prevent leakage, it provided only textual judgments, leading to evidentiary blind spots in disputes. To address these limitations, this paper proposes a novel privacy-preserving perception framework based on the AI Flow paradigm and a collaborative edge-cloud architecture. By deploying a visual desensitizer at the edge, raw images are transformed in real time into abstract feature vectors through nonlinear mapping and stochastic noise injection under the Information Bottleneck principle, ensuring identity-sensitive information is stripped and original images are mathematically unreconstructable. The abstract representations are transmitted to the cloud for behavior recognition and semantic reconstruction via a "dynamic contour" visual language, achieving a critical balance between perception and privacy while enabling illustrative visual reference without exposing raw images.

cs.CV↗

arXiv2026

Beyond Static Artifacts: A Forensic Benchmark for Video Deepfake Reasoning in Vision Language Models

Current Vision-Language Models (VLMs) for deepfake detection excel at identifying spatial artifacts but overlook a critical dimension: temporal inconsistencies in video forgeries. Adapting VLMs to reason about these dynamic cues remains a distinct challenge. To bridge this gap, we propose Forensic Answer-Questioning (FAQ), a large-scale benchmark that formulates temporal deepfake analysis as a multiple-choice task. FAQ introduces a three-level hierarchy to progressively evaluate and equip VLMs with forensic capabilities: (1) Facial Perception, testing the ability to identify static visual artifacts; (2) Temporal Deepfake Grounding, requiring the localization of dynamic forgery artifacts across frames; and (3) Forensic Reasoning, challenging models to synthesize evidence for final authenticity verdicts. We evaluate a range of VLMs on FAQ and generate a corresponding instruction-tuning set, FAQ-IT. Extensive experiments show that models fine-tuned on FAQ-IT achieve advanced performance on both in-domain and cross-dataset detection benchmarks. Ablation studies further validate the impact of our key design choices, confirming that FAQ is the driving force behind the temporal reasoning capabilities of these VLMs.

cs.CV↗

arXiv2026

A Real-Time Privacy-Preserving Behavior Recognition System via Edge-Cloud Collaboration

As intelligent sensing expands into high-privacy environments such as restrooms and changing rooms, the field faces a critical privacy-security paradox. Traditional RGB surveillance raises significant concerns regarding visual recording and storage, while existing privacy-preserving methods-ranging from physical desensitization to traditional cryptographic or obfuscation techniques-often compromise semantic understanding capabilities or fail to guarantee mathematical irreversibility against reconstruction attacks. To address these challenges, this study presents a novel privacy-preserving perception technology based on the AI Flow theoretical framework and an edge-cloud collaborative architecture. The proposed methodology integrates source desensitization with irreversible feature mapping. Leveraging Information Bottleneck theory, the edge device performs millisecond-level processing to transform raw imagery into abstract feature vectors via non-linear mapping and stochastic noise injection. This process constructs a unidirectional information flow that strips identity-sensitive attributes, rendering the reconstruction of original images impossible. Subsequently, the cloud platform utilizes multimodal family models to perform joint inference solely on these abstract vectors to detect abnormal behaviors. This approach fundamentally severs the path to privacy leakage at the architectural level, achieving a breakthrough from video surveillance to de-identified behavior perception and offering a robust solution for risk management in high-sensitivity public spaces.

cs.CR↗

arXiv2026

The Law of Multi-Model Collaboration: Scaling Limits of Model Ensembling for Large Language Models

Recent advances in large language models (LLMs) have been largely driven by scaling laws for individual models, which predict performance improvements as model parameters and data volume increase. However, the capabilities of any single LLM are inherently bounded. One solution originates from intricate interactions among multiple LLMs, rendering their collective performance surpasses that of any constituent model. Despite the rapid proliferation of multi-model integration techniques such as model routing and post-hoc ensembling, a unifying theoretical framework of performance scaling for multi-model collaboration remains absent. In this work, we propose the Law of Multi-model Collaboration, a scaling law that predicts the performance limits of LLM ensembles based on their aggregated parameter budget. To quantify the intrinsic upper bound of multi-model collaboration, we adopt a method-agnostic formulation and assume an idealized integration oracle where the total cross-entropy loss of each sample is determined by the minimum loss of any model in the model pool. Experimental results reveal that multi-model systems follow a power-law scaling with respect to the total parameter count, exhibiting a more significant improvement trend and a lower theoretical loss floor compared to single model scaling. Moreover, ensembles of heterogeneous model families achieve better performance scaling than those formed within a single model family, indicating that model diversity is a primary driver of collaboration gains. These findings suggest that model collaboration represents a critical axis for extending the intelligence frontier of LLMs.

cs.LG↗

arXiv2025

Integrated Ring-based Quantum Key Distribution with Weak Measurement Enhanced Fiber-Optic Sensing Disturbance Magnitude and Location

The deep integration of quantum communication and fiber-optic sensing is pivotal for the development of next-generation multifunctional and highly reliable secure information infrastructure. Here, we present a Sagnac-loop integrated system (SLIS) that, for the first time, combines ring-based quantum key distribution (QKD) with fiber-based weak measurement (WM) enhanced sensing and disturbance localization capabilities. In the event of communication interruption due to external disturbances, the SLIS seamlessly switches to perception system, employing interference measurement and WM techniques to monitor channel disturbances. By integrating null-frequencies localization (NFL) mode, the system precisely determines the disturbance location, enabling rapid identification of security vulnerabilities along the link. Experimental results demonstrate that, over a 30 km Sagnac loop channel, the SLIS achieves a raw key generation rate of 22.4 kbps with stable operation and clear scalability toward network expansion. In terms of perception performance, the SLIS exhibits strong capability for both dynamic and quasi-static disturbances. For dynamic perturbations, the system detects transient impacts and PZT-driven frequency variations down to 100 Hz, and enables long-distance localization via NFL alignment, with improved localization performance as the disturbance position moves farther away along the loop. For quasi-static disturbances, gravitational changes as small as 100 g are resolved, corresponding to a time-delay variation of 9.81 as. This work provides a novel technical pathway toward self-diagnosing, robust quantum networks through integrated communication and sensing functionalities.

quant-ph↗

arXiv2025

See in Depth: Training-Free Surgical Scene Segmentation with Monocular Depth Priors

Pixel-wise segmentation of laparoscopic scenes is essential for computer-assisted surgery but difficult to scale due to the high cost of dense annotations. We propose depth-guided surgical scene segmentation (DepSeg), a training-free framework that utilizes monocular depth as a geometric prior together with pretrained vision foundation models. DepSeg first estimates a relative depth map with a pretrained monocular depth estimation network and proposes depth-guided point prompts, which SAM2 converts into class-agnostic masks. Each mask is then described by a pooled pretrained visual feature and classified via template matching against a template bank built from annotated frames. On the CholecSeg8k dataset, DepSeg improves over a direct SAM2 auto segmentation baseline (35.9% vs. 14.7% mIoU) and maintains competitive performance even when using only 10--20% of the object templates. These results show that depth-guided prompting and template-based classification offer an annotation-efficient segmentation approach.

cs.CV↗

arXiv2025

Systematic Evaluation and Guidelines for Segment Anything Model in Surgical Video Analysis

Surgical video segmentation is critical for AI to interpret spatial-temporal dynamics in surgery, yet model performance is constrained by limited annotated data. The SAM2 model, pretrained on natural videos, offers potential for zero-shot surgical segmentation, but its applicability in complex surgical environments, with challenges like tissue deformation and instrument variability, remains unexplored. We present the first comprehensive evaluation of the zero-shot capability of SAM2 in 9 surgical datasets (17 surgery types), covering laparoscopic, endoscopic, and robotic procedures. We analyze various prompting (points, boxes, mask) and {finetuning (dense, sparse) strategies}, robustness to surgical challenges, and generalization across procedures and anatomies. Key findings reveal that while SAM2 demonstrates notable zero-shot adaptability in structured scenarios (e.g., instrument segmentation, {multi-organ segmentation}, and scene segmentation), its performance varies under dynamic surgical conditions, highlighting gaps in handling temporal coherence and domain-specific artifacts. These results highlight future pathways to adaptive data-efficient solutions for the surgical data science field.

cs.CV↗

arXiv2025

Toward fast, accurate and robust AI prediction of ground states in rotating BEC

We propose an unsupervised deep learning approach for computing the ground state (GS) of rotating Bose-Einstein condensation. To minimize the energy under a mass constraint, our approach introduces two key and novel ingredients: a normalized loss function that exactly enforces the mass constraint, and a training strategy named virtual rotation acceleration that is essential for avoiding local minima and guiding the learning process to the correct quantized vortex phase. Extensive numerical experiments demonstrate the proposed approach as an effective and accurate method to predict GS across physical conditions--from slow to fast rotation and from isotropic to anisotropic confinement. Through further distillation, we establish a unified operator network capable of efficiently generalizing physical parameters across different phases. It enables rapid GS predictions while correctly capturing phase transitions and is applied for inverse problems.

cond-mat.quant-gas↗

arXiv2025

Task-Oriented Feature Compression for Multimodal Understanding via Device-Edge Co-Inference

With the rapid development of large multimodal models (LMMs), multimodal understanding applications are emerging. As most LMM inference requests originate from edge devices with limited computational capabilities, the predominant inference pipeline involves directly forwarding the input data to an edge server which handles all computations. However, this approach introduces high transmission latency due to limited uplink bandwidth of edge devices and significant computation latency caused by the prohibitive number of visual tokens, thus hindering delay-sensitive tasks and degrading user experience. To address this challenge, we propose a task-oriented feature compression (TOFC) method for multimodal understanding in a device-edge co-inference framework, where visual features are merged by clustering and encoded by a learnable and selective entropy model before feature projection. Specifically, we employ density peaks clustering based on K nearest neighbors to reduce the number of visual features, thereby minimizing both data transmission and computational complexity. Subsequently, a learnable entropy model with hyperprior is utilized to encode and decode merged features, further reducing transmission overhead. To enhance compression efficiency, multiple entropy models are adaptively selected based on the characteristics of the visual features, enabling a more accurate estimation of the probability distribution. Comprehensive experiments on seven visual question answering benchmarks validate the effectiveness of the proposed TOFC method. Results show that TOFC achieves up to 52% reduction in data transmission overhead and 63% reduction in system latency while maintaining identical task performance, compared with neural compression ELIC.

eess.SP↗

arXiv2025

AI Flow: Perspectives, Scenarios, and Approaches

Pioneered by the foundational information theory by Claude Shannon and the visionary framework of machine intelligence by Alan Turing, the convergent evolution of information and communication technologies (IT/CT) has created an unbroken wave of connectivity and computation. This synergy has sparked a technological revolution, now reaching its peak with large artificial intelligence (AI) models that are reshaping industries and redefining human-machine collaboration. However, the realization of ubiquitous intelligence faces considerable challenges due to substantial resource consumption in large models and high communication bandwidth demands. To address these challenges, AI Flow has been introduced as a multidisciplinary framework that integrates cutting-edge IT and CT advancements, with a particular emphasis on the following three key points. First, device-edge-cloud framework serves as the foundation, which integrates end devices, edge servers, and cloud clusters to optimize scalability and efficiency for low-latency model inference. Second, we introduce the concept of familial models, which refers to a series of different-sized models with aligned hidden features, enabling effective collaboration and the flexibility to adapt to varying resource constraints and dynamic scenarios. Third, connectivity- and interaction-based intelligence emergence is a novel paradigm of AI Flow. By leveraging communication networks to enhance connectivity, the collaboration among AI models across heterogeneous nodes achieves emergent intelligence that surpasses the capability of any single model. The innovations of AI Flow provide enhanced intelligence, timely responsiveness, and ubiquitous accessibility to AI services, paving the way for the tighter fusion of AI techniques and communication systems.

cs.AI↗

arXiv2025

LCDS: A Logic-Controlled Discharge Summary Generation System Supporting Source Attribution and Expert Review

Despite the remarkable performance of Large Language Models (LLMs) in automated discharge summary generation, they still suffer from hallucination issues, such as generating inaccurate content or fabricating information without valid sources. In addition, electronic medical records (EMRs) typically consist of long-form data, making it challenging for LLMs to attribute the generated content to the sources. To address these challenges, we propose LCDS, a Logic-Controlled Discharge Summary generation system. LCDS constructs a source mapping table by calculating textual similarity between EMRs and discharge summaries to constrain the scope of summarized content. Moreover, LCDS incorporates a comprehensive set of logical rules, enabling it to generate more reliable silver discharge summaries tailored to different clinical fields. Furthermore, LCDS supports source attribution for generated content, allowing experts to efficiently review, provide feedback, and rectify errors. The resulting golden discharge summaries are subsequently recorded for incremental fine-tuning of LLMs. Our project and demo video are in the GitHub repository https://github.com/ycycyc02/LCDS.

cs.CL↗

arXiv2025

Multi-Reference and Adaptive Nonlinear Transform Source-Channel Coding for Wireless Image Semantic Transmission

We propose a multi-reference and adaptive nonlinear transform source-channel coding (MA-NTSCC) system for wireless image semantic transmission to improve rate-distortion (RD) performance by introducing multi-dimensional contexts into the entropy model of the state-of-the-art (SOTA) NTSCC system. Improvements in RD performance of the proposed MA-NTSCC system are particularly significant in high-resolution image transmission under low bandwidth constraints. The proposed multi-reference entropy model leverages correlations within the latent representation in both spatial and channel dimensions. In the spatial dimension, the latent representation is divided into anchors and non-anchors in a checkerboard pattern, where anchors serve as reference to estimate the mutual information between anchors and non-anchors. In the channel dimension, the latent representation is partitioned into multiple groups, and features in previous groups are analyzed to estimate the mutual information between features in previous and current groups. Taking mutual information into account, the entropy model provides an accurate estimation on the entropy, which enables efficient bandwidth allocation and enhances RD performance. Additionally, the proposed lightweight adaptation modules enable the proposed MA-NTSCC model to achieve transmission quality comparable to separately trained models across various channel conditions and bandwidth requirements. In contrast, traditional NTSCC models provide signal-to-noise ratio (SNR)-distortion performance degrading with channel quality deviating from the fixed training SNR, and consume inflexible bandwidth to transmit an image. Comprehensive experiments are conducted to verify the peak signal-to-noise ratio (PSNR) performance and adaptability of the proposed MA-NTSCC model superior to SOTA methods over both additive white Gaussian noise channel and Rayleigh fading channel.

eess.SP↗

arXiv2025

Seed1.5-VL Technical Report

We present Seed1.5-VL, a vision-language foundation model designed to advance general-purpose multimodal understanding and reasoning. Seed1.5-VL is composed with a 532M-parameter vision encoder and a Mixture-of-Experts (MoE) LLM of 20B active parameters. Despite its relatively compact architecture, it delivers strong performance across a wide spectrum of public VLM benchmarks and internal evaluation suites, achieving the state-of-the-art performance on 38 out of 60 public benchmarks. Moreover, in agent-centric tasks such as GUI control and gameplay, Seed1.5-VL outperforms leading multimodal systems, including OpenAI CUA and Claude 3.7. Beyond visual and video understanding, it also demonstrates strong reasoning abilities, making it particularly effective for multimodal reasoning challenges such as visual puzzles. We believe these capabilities will empower broader applications across diverse tasks. In this report, we mainly provide a comprehensive review of our experiences in building Seed1.5-VL across model design, data construction, and training at various stages, hoping that this report can inspire further research. Seed1.5-VL is now accessible at https://www.volcengine.com/ (Volcano Engine Model ID: doubao-1-5-thinking-vision-pro-250428)

cs.CV↗

arXiv2025

BiXiao: An AI-Based Atmospheric Environment Forecasting Model Using Discontinuous Grids

Currently, the technique of numerical model-based atmospheric environment forecasting has becoming mature, yet traditional numerical prediction methods struggle to balance computational costs and forecast accuracy, facing developmental bottlenecks. Recent advancements in artificial intelligence (AI) offer new solutions for weather prediction. However, most existing AI models do not have atmospheric environmental forecasting capabilities, while those with related functionalities remain constrained by grid-dependent data requirements, thus unable to deliver operationally feasible city-scale atmospheric environment forecasts. Here we introduce 'BiXiao', a novel discontinuous-grid AI model for atmospheric environment forecasting. 'BiXiao' couples meteorological and environmental sub-models to generate predictions using site-specific observational data, completing 72-hour forecasts for six major pollutants across all key cities in the Beijing-Tianjin-Hebei region within 30 seconds. In the comparative experiments, the 'BiXiao' model outperforms mainstream numerical models in both computational efficiency and forecast accuracy. It surpasses CAMS with respect of operational 72-hour forecasting and exceeds WRF-Chem's performance in heavy pollution case predictions. The 'BiXiao' shows potential for nationwide application, providing innovative technical support and new perspectives for China's atmospheric environment forecasting operations.

physics.ao-ph↗

arXiv2025

Temporal Propagation of Asymmetric Feature Pyramid for Surgical Scene Segmentation

Surgical scene segmentation is crucial for robot-assisted laparoscopic surgery understanding. Current approaches face two challenges: (i) static image limitations including ambiguous local feature similarities and fine-grained structural details, and (ii) dynamic video complexities arising from rapid instrument motion and persistent visual occlusions. While existing methods mainly focus on spatial feature extraction, they fundamentally overlook temporal dependencies in surgical video streams. To address this, we present temporal asymmetric feature propagation network, a bidirectional attention architecture enabling cross-frame feature propagation. The proposed method contains a temporal query propagator that integrates multi-directional consistency constraints to enhance frame-specific feature representation, and an aggregated asymmetric feature pyramid module that preserves discriminative features for anatomical structures and surgical instruments. Our framework uniquely enables both temporal guidance and contextual reasoning for surgical scene understanding. Comprehensive evaluations on two public benchmarks show the proposed method outperforms the current SOTA methods by a large margin, with +16.4\% mIoU on EndoVis2018 and +3.3\% mAP on Endoscapes2023. The code will be publicly available after paper acceptance.

cs.CV↗

arXiv2025

It Takes Three to Ceilidh: Pension System and Multidimensional Poverty Mitigation in China

This research, employing the Alkire-Foster approach to uncover multidimensional poverty between 2012 and 2020 in China, models and examines the sustainable effects and mechanisms of the three-pillar pension system in household poverty mitigation with the China Family Panel Studies data. The results evince that more participation in the pension system mitigates the probability of being trapped in multidimensional poverty. The findings reveal the significance of synchronising state social insurance, enterprise annuity, and individual commercial insurance. The mitigation effect of market-oriented pillars is achieved through more investment in and consumption for livelihood assets. Based upon the sustainable livelihoods framework, livelihood assets ameliorate household capabilities in human, natural, financial, and psychological capital against exogenous shocks and uncertainties. Our research contributes to the theory and praxis of the facilitating state in new structural economics illustrated by the multi-pillar pension system in sustainable micro-household poverty mitigation.

econ.GN↗

arXiv2024

A weighted scalar auxiliary variable method for solving gradient flows: bridging the nonlinear energy-based and Lagrange multiplier approaches

Two primary scalar auxiliary variable (SAV) approaches are widely applied for simulating gradient flow systems, i.e., the nonlinear energy-based approach and the Lagrange multiplier approach. The former guarantees unconditional energy stability through a modified energy formulation, whereas the latter preserves original energy stability but requires small time steps for numerical solutions. In this paper, we introduce a novel weighted SAV method which integrates these two approaches for the first time. Our method leverages the advantages of both approaches: (i) it ensures the existence of numerical solutions for any time step size with a sufficiently large weight coefficient; (ii) by using a weight coefficient smaller than one, it achieves a discrete energy closer to the original, potentially ensuring stability under mild conditions; and (iii) it maintains consistency in computational cost by utilizing the same time/spatial discretization formulas. We present several theorems and numerical experiments to validate the accuracy, energy stability and superiority of our proposed method.

math.NA↗

arXiv2024

Surgical Scene Segmentation by Transformer With Asymmetric Feature Enhancement

Surgical scene segmentation is a fundamental task for robotic-assisted laparoscopic surgery understanding. It often contains various anatomical structures and surgical instruments, where similar local textures and fine-grained structures make the segmentation a difficult task. Vision-specific transformer method is a promising way for surgical scene understanding. However, there are still two main challenges. Firstly, the absence of inner-patch information fusion leads to poor segmentation performance. Secondly, the specific characteristics of anatomy and instruments are not specifically modeled. To tackle the above challenges, we propose a novel Transformer-based framework with an Asymmetric Feature Enhancement module (TAFE), which enhances local information and then actively fuses the improved feature pyramid into the embeddings from transformer encoders by a multi-scale interaction attention strategy. The proposed method outperforms the SOTA methods in several different surgical segmentation tasks and additionally proves its ability of fine-grained structure recognition. Code is available at https://github.com/cyuan-sjtu/ViT-asym.

cs.CV↗