arXiv subjects

Qianqian Zhang

Publications and source records attributed to Qianqian Zhang.

At least 19 recordsLinked to original sources

arXiv2026

TRACE: Temporal Audit and Condition-aware Evaluation of Streaming Video Understanding

Streaming video understanding requires models to interpret evidence as it arrives, yet current evaluations often report task scores without specifying when evidence becomes valid, how visual history is maintained, or how responses are triggered. As a result, similar scores may correspond to different workloads, failure modes, and operational behavior. We introduce TRACE (Temporal Audit and Condition-aware Evaluation), a condition-aware benchmark and evaluation framework that makes these factors explicit. TRACE combines temporally audited visual tasks with evidence timing and instruction-dependent trigger annotations, a unified causal Core--Adapter protocol that controls information availability while recording actual history processing and response events, and multidimensional reporting of answer quality, timeliness, response-selection behavior, workload, completion, and reliability. On 1,240 records from 517 videos, we evaluate eight publicly available models or systems in eight configurations. We find that nearly identical QA accuracy can mask substantial differences in completion, answer validity, and generation workload, while proactive performance separates into response quality, response delay, false alarms (responses emitted while no target window is currently valid and a later one remains), and missed target windows. These results show that streaming-video performance should be interpreted as execution-conditioned system behavior rather than a single score. Our benchmark and code can be accessed at \href{https://github.com/om-ai-lab/trace-bench}{https://github.com/om-ai-lab/trace-bench}.

cs.CL↗

arXiv2026

VLX-VR: An Agentic-Aware Video Reasoning Model

Real-world video understanding requires integrating visual, audio, textual, and temporal evidence distributed across a video. Yet many pipelines use a fixed video context and single-pass inference, limiting adaptive evidence acquisition when observations are incomplete, ambiguous, or conflicting. We present VLX-VR, an agentic-aware video reasoning model trained within a video reasoning framework defined by a Think--Memory--Observation loop. At each step, VLX-VR determines the needed evidence, invokes read_memory or write_memory, incorporates the returned Observation, and decides whether to continue or produce the task output. We train VLX-VR with multimodal data, including videos and agent trajectories, using reinforcement learning to learn evidence acquisition, memory use, and termination. On MINERVA, VLX-VR achieves state-of-the-art performance among the models included in our comparison, with 78.79% accuracy. Under the original three duration groups, its accuracies are 76.70%, 78.73%, and 80.92%, with a cross-duration accuracy variance of 2.97~$\mathrm{pp}^2$. On correctly answered samples, 96.20% of VLX-VR's reasoning traces are consistent with the MINERVA reference reasoning traces and the evidence described by them, while approximately 75.80% of all evaluated samples satisfy both answer correctness and this evidence-grounded trace criterion. These results show strong performance and broadly stable behavior across durations, while counting, state changes, causal reasoning, and spatial perception remain challenging.

cs.CL↗

arXiv2026

Privacy Preserving Semantic Communications in Wireless Edge Networks with Vision Language Models

Semantic communication has emerged as a promising paradigm for next-generation wireless systems by transmitting high-level semantic features rather than raw bits. However, collaborative devices and multimodal transmission increase privacy risks because sensitive information may leak through inter-device semantic fusion and cross-modal representations. To address this issue, we propose a privacy-preserving semantic communication framework for wireless edge networks. Leveraging a vision-language model (VLM), the framework extracts textual semantics from images and identifies privacy-sensitive entities using a privacy database maintained only at the edge server. Before image transmission, each device removes the identified private regions while preserving useful semantic content. The server then reconstructs the removed regions from the received masked images using textual embeddings and VLM-based semantic priors. To protect textual information, we design an encrypted semantic-channel transceiver using physical-layer keys generated from reciprocal wireless channels, without pre-shared keys. We also introduce a semantic information bottleneck to suppress redundant information across multiple devices. The framework is evaluated against a strong model-aware adversary that can intercept wireless transmissions and access edge-device model parameters but not server-side data. Simulation results show that the proposed method reduces privacy leakage by more than 50% compared with a semantic communication scheme without privacy protection, while the authorized server achieves a 48% improvement in perceptual reconstruction quality over the adversary. The estimated mutual information between transmitted representations approaches 0 bit, indicating effective suppression of cross-device semantic redundancy.

cs.CR↗

arXiv2026

CRB-Driven Beamforming and Trajectory Optimization for UAV-assisted ISAC System

In this paper, we study an unmanned aerial vehicle (UAV)-assisted integrated sensing and communication (ISAC) system, where a UAV enhances the sensing capability of a base station (BS) towards a target while ensuring reliable communication towards a downlink user. This architecture is practically attractive for future wireless networks due to the UAV's controllable mobility and adaptive sensing coverage in wireless environments. The sensing performance is characterized by the average Cramér-Rao bound (CRB), which quantifies the minimum variance of the unbiased angle-of-arrival estimation. To enhance the sensing performance, the UAV trajectory and beamforming parameters are jointly optimized under power and mobility constraints, while satisfying communication requirements to the downlink user. To address the resulting non-convex problem, we employ null-space projection for beamforming design and adopt deep reinforcement learning for the trajectory optimization over a discrete-time scale. In each time slot, beamforming is optimized based on the channel state information to improve CRB performance while mitigating interference between the BS and the communication user. Simulation results demonstrate that the proposed method significantly reduces the time-averaged CRB by over 10%, compared with the ISAC system without UAV assistance, and also achieves a higher sensing accuracy than both the fixed-UAV-trajectory and the maximum-ratio-transmission-based beamforming benchmarks.

cs.IT↗

arXiv2026

SilverTorch: A Unified Model-based System to Democratize Large-Scale Recommendation on GPUs

Serving deep learning based recommendation models (DLRM) at scale is challenging. Existing approaches rely on dedicated ANN indexing and filtering services on CPUs, suffering from non-negligible costs and missing co-design opportunities. Such inefficiency makes them difficult to support complex model architectures, such as learned similarities and multi-task retrieval. In this paper, we present SilverTorch, a model-based serving system that brings all components into one unified model. It unifies model serving by replacing standalone indexing and filtering services with model layers. We propose a model-based GPU Bloom index for feature filtering and a fused Int8 ANN kernel for nearest neighbor search. Through co-design of the ANN search and feature filtering, we reduce GPU memory usage and eliminate computation. Benefiting from this design, we scale up retrieval by introducing an OverArch scoring layer and a multi-task retrieval with a Value Model to aggregate scores. These advancements improve the retrieval accuracy and enable future studies for serving more complex models. Our evaluation on industry-scale datasets show that SilverTorch achieves up to 23.7\times higher throughput compared to the state-of-the-art approaches. We also demonstrate that SilverTorch solution is 13.35\times more cost-efficient than CPU-based solution while improving accuracy via serving more complex models. SilverTorch is deployed at scale, serving hundreds of models online and supporting recommendation for diverse applications.

cs.IR↗

arXiv2026

NTIRE 2026 The 3rd Restore Any Image Model (RAIM) Challenge: Professional Image Quality Assessment (Track 1)

In this paper, we present an overview of the NTIRE 2026 challenge on the 3rd Restore Any Image Model in the Wild, specifically focusing on Track 1: Professional Image Quality Assessment. Conventional Image Quality Assessment (IQA) typically relies on scalar scores. By compressing complex visual characteristics into a single number, these methods fundamentally struggle to distinguish subtle differences among uniformly high-quality images. Furthermore, they fail to articulate why one image is superior, lacking the reasoning capabilities required to provide guidance for vision tasks. To bridge this gap, recent advancements in Multimodal Large Language Models (MLLMs) offer a promising paradigm. Inspired by this potential, our challenge establishes a novel benchmark exploring the ability of MLLMs to mimic human expert cognition in evaluating high-quality image pairs. Participants were tasked with overcoming critical bottlenecks in professional scenarios, centering on two primary objectives: (1) Comparative Quality Selection: reliably identifying the visually superior image within a high-quality pair; and (2) Interpretative Reasoning: generating grounded, expert-level explanations that detail the rationale behind the selection. In total, the challenge attracted nearly 200 registrations and over 2,500 submissions. The top-performing methods significantly advanced the state of the art in professional IQA. The challenge dataset is available at https://github.com/narthchin/RAIM-PIQA, and the official homepage is accessible at https://www.codabench.org/competitions/12789/.

cs.CV↗

arXiv2026

On the Capacity Region of Additive-Multiplicative MAC with Heterogeneous Input Constraints

This paper characterizes the capacity region of a two-user additive-multiplicative multiple access channel (AM-MAC) under heterogeneous input constraints. This model captures the fundamental limits of symbiotic radio, where an active primary transmitter (PT) conveys information via active transmission subject to an average power constraint, while a passive backscatter device (BD) modulates signals through backscattering under a peak amplitude constraint. Our main results are threefold. Firstly, we prove that the sum-rate capacity equals the PT's point-to-point capacity, achieved when the PT employs Gaussian signaling and the BD acts as a pure reflector to assist the PT's transmission. Secondly, to achieve the BD's maximum achievable rate, the PT must adopt a constant-envelope signaling strategy, while the optimal BD distribution exhibits a concentric-circle structure with a uniform phase. Thirdly, for the remaining boundary points, we establish that the optimal PT signal consists of a continuous uniform phase and a discrete amplitude, whereas the optimal BD distribution is fully discrete. Finally, numerical results are provided to characterized the capacity region by solving a specialized nonlinear optimization problem. To demonstrate the practical implications, we also characterize an baseline rate pair and evaluate the overall performance of the AM-MAC.

cs.IT↗

arXiv2026

DLRMamba: Distilling Low-Rank Mamba for Edge Multispectral Fusion Object Detection

Multispectral fusion object detection is a critical task for edge-based maritime surveillance and remote sensing, demanding both high inference efficiency and robust feature representation for high-resolution inputs. However, current State Space Models (SSMs) like Mamba suffer from significant parameter redundancy in their standard 2D Selective Scan (SS2D) blocks, which hinders deployment on resource-constrained hardware and leads to the loss of fine-grained structural information during conventional compression. To address these challenges, we propose the Low-Rank Two-Dimensional Selective Structured State Space Model (Low-Rank SS2D), which reformulates state transitions via matrix factorization to exploit intrinsic feature sparsity. Furthermore, we introduce a Structure-Aware Distillation strategy that aligns the internal latent state dynamics of the student with a full-rank teacher model to compensate for potential representation degradation. This approach substantially reduces computational complexity and memory footprint while preserving the high-fidelity spatial modeling required for object recognition. Extensive experiments on five benchmark datasets and real-world edge platforms, such as Raspberry Pi 5, demonstrate that our method achieves a superior efficiency-accuracy trade-off, significantly outperforming existing lightweight architectures in practical deployment scenarios.

cs.CV↗

arXiv2026

Small Target Detection Based on Mask-Enhanced Attention Fusion of Visible and Infrared Remote Sensing Images

Targets in remote sensing images are usually small, weakly textured, and easily disturbed by complex backgrounds, challenging high-precision detection with general algorithms. Building on our earlier ESM-YOLO, this work presents ESM-YOLO+ as a lightweight visible infrared fusion network. To enhance detection, ESM-YOLO+ includes two key innovations. (1) A Mask-Enhanced Attention Fusion (MEAF) module fuses features at the pixel level via learnable spatial masks and spatial attention, effectively aligning RGB and infrared features, enhancing small-target representation, and alleviating cross-modal misalignment and scale heterogeneity. (2) Training-time Structural Representation (SR) enhancement provides auxiliary supervision to preserve fine-grained spatial structures during training, boosting feature discriminability without extra inference cost. Extensive experiments on the VEDAI and DroneVehicle datasets validate ESM-YOLO+'s superiority. The model achieves 84.71\% mAP on VEDAI and 74.0\% mAP on DroneVehicle, while greatly reducing model complexity, with 93.6\% fewer parameters and 68.0\% lower GFLOPs than the baseline. These results confirm that ESM-YOLO+ integrates strong performance with practicality for real-time deployment, providing an effective solution for high-performance small-target detection in complex remote sensing scenes.

cs.CV↗

arXiv2026

Reflections over the Sea: Reconfigurable Intelligent Surface for Maritime Self-Powered Communications

Maritime communication is becoming a vital component of 6G networks, driven by the rapid expansion of the maritime economy. However, existing technologies face critical challenges in signal coverage, availability, and robustness, especially under harsh sea conditions. This paper proposes a novel framework for the maritime Internet-of-Things (IoT) communications that leverages the reconfigurable intelligent surface (RIS) mounted on offshore infrastructures, such as wind turbines, to enhance coverage and reliability. To capture dynamic maritime environment, a near-ocean-surface channel model is developed considering the impact of sea waves. In addition, a wave energy harvesting (EH) system is designed to self-power IoT sensors for data acquisition, processing, and transmission. To support real-time adaptation, channel state information is continuously measured to optimize RIS reflection parameters and maximize multi-user communication rates. Simulation results show that the proposed system significantly improves IoT communication performance by over 20%, under harsh sea conditions.

cs.IT↗

arXiv2025

Privacy Preserving Semantic Communications Using Vision Language Models: A Segmentation and Generation Approach

Semantic communication has emerged as a promising paradigm for next-generation wireless systems, improving the communication efficiency by transmitting high-level semantic features. However, reliance on unimodal representations can degrade reconstruction under poor channel conditions, and privacy concerns of the semantic information attack also gain increasing attention. In this work, a privacy-preserving semantic communication framework is proposed to protect sensitive content of the image data. Leveraging a vision-language model (VLM), the proposed framework identifies and removes private content regions from input images prior to transmission. A shared privacy database enables semantic alignment between the transmitter and receiver to ensure consistent identification of sensitive entities. At the receiver, a generative module reconstructs the masked regions using learned semantic priors and conditioned on the received text embedding. Simulation results show that generalizes well to unseen image processing tasks, improves reconstruction quality at the authorized receiver by over 10% using text embedding, and reduces identity leakage to the eavesdropper by more than 50%.

eess.SP↗

arXiv2025

Unifying Language Agent Algorithms with Graph-based Orchestration Engine for Reproducible Agent Research

Language agents powered by large language models (LLMs) have demonstrated remarkable capabilities in understanding, reasoning, and executing complex tasks. However, developing robust agents presents significant challenges: substantial engineering overhead, lack of standardized components, and insufficient evaluation frameworks for fair comparison. We introduce Agent Graph-based Orchestration for Reasoning and Assessment (AGORA), a flexible and extensible framework that addresses these challenges through three key contributions: (1) a modular architecture with a graph-based workflow engine, efficient memory management, and clean component abstraction; (2) a comprehensive suite of reusable agent algorithms implementing state-of-the-art reasoning approaches; and (3) a rigorous evaluation framework enabling systematic comparison across multiple dimensions. Through extensive experiments on mathematical reasoning and multimodal tasks, we evaluate various agent algorithms across different LLMs, revealing important insights about their relative strengths and applicability. Our results demonstrate that while sophisticated reasoning approaches can enhance agent capabilities, simpler methods like Chain-of-Thought often exhibit robust performance with significantly lower computational overhead. AGORA not only simplifies language agent development but also establishes a foundation for reproducible agent research through standardized evaluation protocols.

cs.CL↗

arXiv2025

Selective Structured State Space for Multispectral-fused Small Target Detection

Target detection in high-resolution remote sensing imagery faces challenges due to the low recognition accuracy of small targets and high computational costs. The computational complexity of the Transformer architecture increases quadratically with image resolution, while Convolutional Neural Networks (CNN) architectures are forced to stack deeper convolutional layers to expand their receptive fields, leading to an explosive growth in computational demands. To address these computational constraints, we leverage Mamba's linear complexity for efficiency. However, Mamba's performance declines for small targets, primarily because small targets occupy a limited area in the image and have limited semantic information. Accurate identification of these small targets necessitates not only Mamba's global attention capabilities but also the precise capture of fine local details. To this end, we enhance Mamba by developing the Enhanced Small Target Detection (ESTD) module and the Convolutional Attention Residual Gate (CARG) module. The ESTD module bolsters local attention to capture fine-grained details, while the CARG module, built upon Mamba, emphasizes spatial and channel-wise information, collectively improving the model's ability to capture distinctive representations of small targets. Additionally, to highlight the semantic representation of small targets, we design a Mask Enhanced Pixel-level Fusion (MEPF) module for multispectral fusion, which enhances target features by effectively fusing visible and infrared multimodal information.

cs.CV↗

arXiv2025

VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model

Recently DeepSeek R1 has shown that reinforcement learning (RL) can substantially improve the reasoning capabilities of Large Language Models (LLMs) through a simple yet effective design. The core of R1 lies in its rule-based reward formulation, which leverages tasks with deterministic ground-truth answers to enable precise and stable reward computation. In the visual domain, we similarly observe that a wide range of visual understanding tasks are inherently equipped with well-defined ground-truth annotations. This property makes them naturally compatible with rule-based reward mechanisms. Motivated by this observation, we investigate the extension of R1-style reinforcement learning to Vision-Language Models (VLMs), aiming to enhance their visual reasoning capabilities. To this end, we develop VLM-R1, a dedicated framework designed to harness RL for improving VLMs' performance on general vision-language tasks. Using this framework, we further explore the feasibility of applying RL to visual domain. Experimental results indicate that the RL-based model not only delivers competitive performance on visual understanding tasks but also surpasses Supervised Fine-Tuning (SFT) in generalization ability. Furthermore, we conduct comprehensive ablation studies that uncover a series of noteworthy insights, including the presence of reward hacking in object detection, the emergence of the "OD aha moment", the impact of training data quality, and the scaling behavior of RL across different model sizes. Through these analyses, we aim to deepen the understanding of how reinforcement learning enhances the capabilities of vision-language models, and we hope our findings and open-source contributions will support continued progress in the vision-language RL community. Our code and model are available at https://github.com/om-ai-lab/VLM-R1

cs.CV↗

arXiv2024

Latency Optimization in LEO Satellite Communications with Hybrid Beam Pattern and Interference Control

The rapid advancement of low Earth orbit (LEO) satellite communication systems has significantly enhanced global connectivity, offering high-capacity, low-latency services crucial for next-generation applications. However, the dense configuration of LEO constellations poses challenges in resource allocation optimization and interference management, complicating coexistence with other communication systems. To address these limitations, this paper proposes a novel framework for optimizing the beam scheduling and resource allocation in multi-beam LEO systems. To satisfy the uneven terrestrial traffic demand, a hybrid beam pattern is employed to enhance the downlink quality of service and minimize the transmission latency from LEO satellites to ground user terminals. Additionally, a dynamic co-channel interference (CCI) control mechanism is developed to mitigate inter-beam interference within the LEO constellation and limit cross-system interference affecting protected users from other networks. The problem of user-beam-frequency allocation with power optimization is formulated as a mixed-integer dynamic programming model and solved using a low-complexity neural network-based graph generation algorithm. Simulation results show that the proposed approach outperforms the baseline methods of full frequency reuse and single-channel transmission, and highlights the potential for further performance improvement with multi-user transmissions.

cs.IT↗

arXiv2024

OmChat: A Recipe to Train Multimodal Language Models with Strong Long Context and Video Understanding

We introduce OmChat, a model designed to excel in handling long contexts and video understanding tasks. OmChat's new architecture standardizes how different visual inputs are processed, making it more efficient and adaptable. It uses a dynamic vision encoding process to effectively handle images of various resolutions, capturing fine details across a range of image qualities. OmChat utilizes an active progressive multimodal pretraining strategy, which gradually increases the model's capacity for long contexts and enhances its overall abilities. By selecting high-quality data during training, OmChat learns from the most relevant and informative data points. With support for a context length of up to 512K, OmChat demonstrates promising performance in tasks involving multiple images and videos, outperforming most open-source models in these benchmarks. Additionally, OmChat proposes a prompting strategy for unifying complex multimodal inputs including single image text, multi-image text and videos, and achieving competitive performance on single-image benchmarks. To further evaluate the model's capabilities, we proposed a benchmark dataset named Temporal Visual Needle in a Haystack. This dataset assesses OmChat's ability to comprehend temporal visual details within long videos. Our analysis highlights several key factors contributing to OmChat's success: support for any-aspect high image resolution, the active progressive pretraining strategy, and high-quality supervised fine-tuning datasets. This report provides a detailed overview of OmChat's capabilities and the strategies that enhance its performance in visual understanding.

cs.CV↗

arXiv2024

On the Capacity Region of Reconfigurable Intelligent Surface Assisted Symbiotic Radios

In this paper, we are interested in a reconfigurable intelligent surface (RIS)-assisted symbiotic radio (SR) system, where an RIS assists a primary transmission by passive beamforming and simultaneously acts as an information transmitter by periodically adjusting its reflection coefficients. The above RIS functions innately enable a new multiplicative multiple access channel (M-MAC), where the primary and secondary signals are superposed in a multiplicative manner. To pursue the fundamental performance limits of M-MAC, we focus on the characterization of the capacity region for such a system when the direct link is blocked. Due to the reflection nature of RIS, the signal transmitted from the RIS should satisfy a passive reflection constraint. We consider two types of passive reflection constraints, one for the case that only the phases of the RIS can be adjusted, while the other for the case that both the amplitude and the phase can be adjusted. Under the passive reflection constraints at the RIS as well as the average power constraint at the primary transmitter (PTx), we characterize the capacity region of RIS-assisted SR. It is observed that: 1) the number of sum-rate-optimal points on the boundary of the capacity region is infinite; 2) for the rate pairs with the maximum sum rate, the optimal distribution of the amplitude of the primary signal is a continuous Rayleigh distribution, while for the remaining rate pairs on the capacity region boundary, the optimal amplitude of the primary signal is discrete; 3) the adjustment of the amplitude for the RIS can enlarge the capacity region as compared to the phase-adjusted-only case.

cs.IT↗

arXiv2024

Modulation Design and Optimization for RIS-Assisted Symbiotic Radios

In reconfigurable intelligent surface (RIS)-assisted symbiotic radio (SR), the RIS acts as a secondary transmitter by modulating its information bits over the incident primary signal and simultaneously assists the primary transmission, then a cooperative receiver is used to jointly decode the primary and secondary signals. Most existing works of SR focus on using RIS to enhance the reflecting link while ignoring the ambiguity problem for the joint detection caused by the multiplication relationship of the primary and secondary signals. Particularly, in case of a blocked direct link, joint detection will suffer from severe performance loss due to the ambiguity, when using the conventional on-off keying and binary phase shift keying modulation schemes for RIS. To address this issue, we propose a novel modulation scheme for RIS-assisted SR that divides the phase-shift matrix into two components: the symbol-invariant and symbol-varying components, which are used to assist the primary transmission and carry the secondary signal, respectively. To design these two components, we focus on the detection of the composite signal formed by the primary and secondary signals, through which a problem of minimizing the bit error rate (BER) of the composite signal is formulated to improve both the BER performance of the primary and secondary ones. By solving the problem, we derive the closed-form solution of the optimal symbol-invariant and symbol-varying components, which is related to the channel strength ratio of the direct link to the reflecting link. Moreover, theoretical BER performance is analyzed. Finally, simulation results show the superiority of the proposed modulation scheme over its conventional counterpart.

cs.IT↗