arXiv subjects

Yao Zhu

Publications and source records attributed to Yao Zhu.

At least 55 records · Page 3Linked to original sources

arXiv2025

SOPSeg: Prompt-based Small Object Instance Segmentation in Remote Sensing Imagery

Extracting small objects from remote sensing imagery plays a vital role in various applications, including urban planning, environmental monitoring, and disaster management. While current research primarily focuses on small object detection, instance segmentation for small objects remains underexplored, with no dedicated datasets available. This gap stems from the technical challenges and high costs of pixel-level annotation for small objects. While the Segment Anything Model (SAM) demonstrates impressive zero-shot generalization, its performance on small-object segmentation deteriorates significantly, largely due to the coarse 1/16 feature resolution that causes severe loss of fine spatial details. To this end, we propose SOPSeg, a prompt-based framework specifically designed for small object segmentation in remote sensing imagery. It incorporates a region-adaptive magnification strategy to preserve fine-grained details, and employs a customized decoder that integrates edge prediction and progressive refinement for accurate boundary delineation. Moreover, we introduce a novel prompting mechanism tailored to the oriented bounding boxes widely adopted in remote sensing applications. SOPSeg outperforms existing methods in small object segmentation and facilitates efficient dataset construction for remote sensing tasks. We further construct a comprehensive small object instance segmentation dataset based on SODA-A, and will release both the model and dataset to support future research.

cs.CV↗

arXiv2025

DipSVD: Dual-importance Protected SVD for Efficient LLM Compression

The ever-increasing computational demands and deployment costs of large language models (LLMs) have spurred numerous compressing methods. Compared to quantization and unstructured pruning, SVD compression offers superior hardware compatibility and theoretical guarantees. However, existing SVD-based methods focus on the overall discrepancy between the original and compressed matrices while overlooking the protection of critical components within the matrix, which leads to inferior performance in the compressed models. This paper proposes a dual-level importance protection mechanism to enhance SVD-based compression methods: (1) local importance protection: preserving the most critical singular vectors within each weight matrix through channel-weighted data whitening; and (2) global importance protection: enabling less important layers to bear a greater portion of the compression burden through either a heuristic or optimization-based approach, thereby minimizing the impact of compression on critical layers. Extensive experiments demonstrate that DipSVD outperforms existing SVD-based compression approaches across multiple benchmarks, achieving superior model performance especially at high model compression ratios.

cs.LG↗

arXiv2025

Towards Understanding How Knowledge Evolves in Large Vision-Language Models

Large Vision-Language Models (LVLMs) are gradually becoming the foundation for many artificial intelligence applications. However, understanding their internal working mechanisms has continued to puzzle researchers, which in turn limits the further enhancement of their capabilities. In this paper, we seek to investigate how multimodal knowledge evolves and eventually induces natural languages in LVLMs. We design a series of novel strategies for analyzing internal knowledge within LVLMs, and delve into the evolution of multimodal knowledge from three levels, including single token probabilities, token probability distributions, and feature encodings. In this process, we identify two key nodes in knowledge evolution: the critical layers and the mutation layers, dividing the evolution process into three stages: rapid evolution, stabilization, and mutation. Our research is the first to reveal the trajectory of knowledge evolution in LVLMs, providing a fresh perspective for understanding their underlying mechanisms. Our codes are available at https://github.com/XIAO4579/Vlm-interpretability.

cs.CV↗

arXiv2025

Efficient Integration of Distributed Learning Services in Next-Generation Wireless Networks

Distributed learning (DL) is considered a cornerstone of intelligence enabler, since it allows for collaborative training without the necessity for local clients to share raw data with other parties, thereby preserving privacy and security. Integrating DL into the 6G networks requires coexistence design with existing services such as high-bandwidth (HB) traffic like eMBB. Current designs in the literature mainly focus on communication round (CR)-wise designs that assume a fixed resource allocation during each CR. However, fixed resource allocation within a CR is a highly inefficient and inaccurate representation of the system's realistic behavior. This is due to the heterogeneous nature of the system, where clients inherently need to access the network at different times. This work zooms into one arbitrary communication round and demonstrates the importance of considering a time-dependent resource-sharing design with HB traffic. We propose a time-dependent optimization problem for minimizing the consumed time and energy by DL within the CR. Due to its intractability, a session-based optimization problem has been proposed assuming a large-scale coherence time. An iterative algorithm has been designed to solve such problems and simulation results confirm the importance of such efficient and accurate integration design.

eess.SY↗

arXiv2025

AdvLogo: Adversarial Patch Attack against Object Detectors based on Diffusion Models

With the rapid development of deep learning, object detectors have demonstrated impressive performance; however, vulnerabilities still exist in certain scenarios. Current research exploring the vulnerabilities using adversarial patches often struggles to balance the trade-off between attack effectiveness and visual quality. To address this problem, we propose a novel framework of patch attack from semantic perspective, which we refer to as AdvLogo. Based on the hypothesis that every semantic space contains an adversarial subspace where images can cause detectors to fail in recognizing objects, we leverage the semantic understanding of the diffusion denoising process and drive the process to adversarial subareas by perturbing the latent and unconditional embeddings at the last timestep. To mitigate the distribution shift that exposes a negative impact on image quality, we apply perturbation to the latent in frequency domain with the Fourier Transform. Experimental results demonstrate that AdvLogo achieves strong attack performance while maintaining high visual quality.

cs.CV↗

arXiv2025

High Dynamic Range Video Compression: A Large-Scale Benchmark Dataset and A Learned Bit-depth Scalable Compression Algorithm

Recently, learned video compression (LVC) is undergoing a period of rapid development. However, due to absence of large and high-quality high dynamic range (HDR) video training data, LVC on HDR video is still unexplored. In this paper, we are the first to collect a large-scale HDR video benchmark dataset, named HDRVD2K, featuring huge quantity, diverse scenes and multiple motion types. HDRVD2K fills gaps of video training data and facilitate the development of LVC on HDR videos. Based on HDRVD2K, we further propose the first learned bit-depth scalable video compression (LBSVC) network for HDR videos by effectively exploiting bit-depth redundancy between videos of multiple dynamic ranges. To achieve this, we first propose a compression-friendly bit-depth enhancement module (BEM) to effectively predict original HDR videos based on compressed tone-mapped low dynamic range (LDR) videos and dynamic range prior, instead of reducing redundancy only through spatio-temporal predictions. Our method greatly improves the reconstruction quality and compression performance on HDR videos. Extensive experiments demonstrate the effectiveness of HDRVD2K on learned HDR video compression and great compression performance of our proposed LBSVC network. Code and dataset will be released in https://github.com/sdkinda/HDR-Learned-Video-Coding.

cs.CV↗

arXiv2025

Physical Layer Deception in OFDM Systems

As a promising technology, physical layer security (PLS) enhances security by leveraging the physical characteristics of communication channels. However, it commonly takes the legitimate user more effort to secure its data, compared to that required by the eavesdropper to intercept the communication. To address this imbalance, we propose a physical layer deception (PLD) framework, which applies random deceptive ciphering combined with orthogonal frequency-division multiplexing (OFDM) to deceive eavesdroppers with falsified information, preventing them from wiretapping. While ensuring the same level of confidentiality as traditional PLS methods, the PLD approach additionally introduces a deception mechanism, which remains effective even when the eavesdropper has the same knowledge about the transmitter as the legitimate receiver. Through detailed theoretical analysis and numerical simulations, we prove the superiority of our method over the conventional PLS approach.

cs.IT↗

arXiv2025

Physical Layer Deception with Non-Orthogonal Multiplexing

Physical layer security (PLS) is a promising technology to secure wireless communications by exploiting the physical properties of the wireless channel. However, the passive nature of PLS creates a significant imbalance between the effort required by eavesdroppers and legitimate users to secure data. To address this imbalance, in this article, we propose a novel framework of physical layer deception (PLD), which combines PLS with deception technologies to actively counteract wiretapping attempts. Combining a two-stage encoder with randomized ciphering and non-orthogonal multiplexing, the PLD approach enables the wireless communication system to proactively counter eavesdroppers with deceptive messages. Relying solely on the superiority of the legitimate channel over the eavesdropping channel, the PLD framework can effectively protect the confidentiality of the transmitted messages, even against eavesdroppers who possess knowledge equivalent to that of the legitimate receiver. We prove the validity of the PLD framework with in-depth analyses and demonstrate its superiority over conventional PLS approaches with comprehensive numerical benchmarks.

cs.CR↗

arXiv2024

An Efficient Framework for Enhancing Discriminative Models via Diffusion Techniques

Image classification serves as the cornerstone of computer vision, traditionally achieved through discriminative models based on deep neural networks. Recent advancements have introduced classification methods derived from generative models, which offer the advantage of zero-shot classification. However, these methods suffer from two main drawbacks: high computational overhead and inferior performance compared to discriminative models. Inspired by the coordinated cognitive processes of rapid-slow pathway interactions in the human brain during visual signal recognition, we propose the Diffusion-Based Discriminative Model Enhancement Framework (DBMEF). This framework seamlessly integrates discriminative and generative models in a training-free manner, leveraging discriminative models for initial predictions and endowing deep neural networks with rethinking capabilities via diffusion models. Consequently, DBMEF can effectively enhance the classification accuracy and generalization capability of discriminative models in a plug-and-play manner. We have conducted extensive experiments across 17 prevalent deep model architectures with different training methods, including both CNN-based models such as ResNet and Transformer-based models like ViT, to demonstrate the effectiveness of the proposed DBMEF. Specifically, the framework yields a 1.51\% performance improvement for ResNet-50 on the ImageNet dataset and 3.02\% on the ImageNet-A dataset. In conclusion, our research introduces a novel paradigm for image classification, demonstrating stable improvements across different datasets and neural networks. The code is available at https://github.com/ChunXiaostudy/DBMEF.

cs.CV↗

arXiv2024

Fairness-aware Age-of-Information Minimization in WPT-Assisted Short-Packet Data Collection for mURLLC

The technological landscape is rapidly evolving toward large-scale systems. Networks supporting massive connectivity through numerous Internet of Things (IoT) devices are at the forefront of this advancement. In this paper, we examine Wireless Power Transfer (WPT)-enabled networks, where a server requires to collect data from these IoT devices to compute a task with massive Ultra-Reliable and Low-Latency Communication (mURLLC) services.} We focus on information freshness, using Age-of-Information (AoI) as the key performance metric. Specifically, we aim to minimize the maximum AoI among IoT devices by optimizing the scheduling policy. Our analytical findings demonstrate the convexity of the problem, enabling efficient solutions. We introduce the concept of AoI-oriented cluster capacity and analyze the relationship between the number of supported devices and network AoI performance. Numerical simulations validate our proposed approach's effectiveness in enhancing AoI performance, highlighting its potential for guiding the design of future IoT systems requiring mURLLC services.

cs.IT↗

arXiv2024

Constrained Coding and Deep Learning Aided Threshold Detection for Resistive Memories

Resistive random access memory (ReRAM) is a promising emerging non-volatile memory (NVM) technology that shows high potential for both data storage and computing. However, its crossbar array architecture leads to the sneak path problem, which may severely degrade the reliability of data stored in the ReRAM cell. Due to the complication of memory physics and unique features of the sneak path induced interference (SPI), it is difficult to derive an accurate channel model for it. The deep learning (DL)-based detection scheme \cite{zhong2020sneakdl} can better mitigate the SPI, at the cost of additional power consumption and read latency. In this letter, we first propose a novel CC scheme which can not only reduce the SPI in the memory array, but also effectively differentiate the memory arrays into two categories of sneak-path-free and sneak-path-affected arrays. For the sneak-path-free arrays, we can use a simple middle-point threshold detector to detect the low and high resistance cells of ReRAM. For the sneak-path-affected arrays, a DL detector is first trained off-line (prior to the data detection of ReRAM). To avoid the additional power consumption and latency introduced by the DL detector, we further propose a DL-based threshold detector, whose detection threshold can be derived based on the outputs of the DL detector. It is then utilized for the online data detection of all the identified sneak-path-affected arrays. Simulation results demonstrate that the above CC and DL aided threshold detection scheme can effectively mitigate the SPI of the ReRAM array and achieve better error rate performance than the prior art detection schemes, without the prior knowledge of the channel.

cs.IT↗

arXiv2024

A Semantic Model for Physical Layer Deception

Physical layer deception (PLD) is a novel security mechanism that combines physical layer security (PLS) with deception technologies to actively defend against eavesdroppers. In this paper, we establish a novel semantic model for PLD that evaluates its performance in terms of semantic distortion. By analyzing semantic distortion at varying levels of knowledge on the receiver's part regarding the key, we derive the receiver's optimal decryption strategy, and consequently, the transmitter's optimal deception strategy. The proposed semantic model provides a more generic understanding of the PLD approach independent from coding or multiplexing schemes, and allows for efficient real-time adaptation to fading channels.

cs.IT↗

arXiv2024

Survey on Knowledge Distillation for Large Language Models: Methods, Evaluation, and Application

Large Language Models (LLMs) have showcased exceptional capabilities in various domains, attracting significant interest from both academia and industry. Despite their impressive performance, the substantial size and computational demands of LLMs pose considerable challenges for practical deployment, particularly in environments with limited resources. The endeavor to compress language models while maintaining their accuracy has become a focal point of research. Among the various methods, knowledge distillation has emerged as an effective technique to enhance inference speed without greatly compromising performance. This paper presents a thorough survey from three aspects: method, evaluation, and application, exploring knowledge distillation techniques tailored specifically for LLMs. Specifically, we divide the methods into white-box KD and black-box KD to better illustrate their differences. Furthermore, we also explored the evaluation tasks and distillation effects between different distillation methods, and proposed directions for future research. Through in-depth understanding of the latest advancements and practical applications, this survey provides valuable resources for researchers, paving the way for sustained progress in this field.

cs.CL↗

arXiv2024

GreenCOD: A Green Camouflaged Object Detection Method

We introduce GreenCOD, a green method for detecting camouflaged objects, distinct in its avoidance of backpropagation techniques. GreenCOD leverages gradient boosting and deep features extracted from pre-trained Deep Neural Networks (DNNs). Traditional camouflaged object detection (COD) approaches often rely on complex deep neural network architectures, seeking performance improvements through backpropagation-based fine-tuning. However, such methods are typically computationally demanding and exhibit only marginal performance variations across different models. This raises the question of whether effective training can be achieved without backpropagation. Addressing this, our work proposes a new paradigm that utilizes gradient boosting for COD. This approach significantly simplifies the model design, resulting in a system that requires fewer parameters and operations and maintains high performance compared to state-of-the-art deep learning models. Remarkably, our models are trained without backpropagation and achieve the best performance with fewer than 20G Multiply-Accumulate Operations (MACs). This new, more efficient paradigm opens avenues for further exploration in green, backpropagation-free model training.

cs.CV↗

arXiv2024

Advancing Out-of-Distribution Detection through Data Purification and Dynamic Activation Function Design

In the dynamic realms of machine learning and deep learning, the robustness and reliability of models are paramount, especially in critical real-world applications. A fundamental challenge in this sphere is managing Out-of-Distribution (OOD) samples, significantly increasing the risks of model misclassification and uncertainty. Our work addresses this challenge by enhancing the detection and management of OOD samples in neural networks. We introduce OOD-R (Out-of-Distribution-Rectified), a meticulously curated collection of open-source datasets with enhanced noise reduction properties. In-Distribution (ID) noise in existing OOD datasets can lead to inaccurate evaluation of detection algorithms. Recognizing this, OOD-R incorporates noise filtering technologies to refine the datasets, ensuring a more accurate and reliable evaluation of OOD detection algorithms. This approach not only improves the overall quality of data but also aids in better distinguishing between OOD and ID samples, resulting in up to a 2.5\% improvement in model accuracy and a minimum 3.2\% reduction in false positives. Furthermore, we present ActFun, an innovative method that fine-tunes the model's response to diverse inputs, thereby improving the stability of feature extraction and minimizing specificity issues. ActFun addresses the common problem of model overconfidence in OOD detection by strategically reducing the influence of hidden units, which enhances the model's capability to estimate OOD uncertainty more accurately. Implementing ActFun in the OOD-R dataset has led to significant performance enhancements, including an 18.42\% increase in AUROC of the GradNorm method and a 16.93\% decrease in FPR95 of the Energy method. Overall, our research not only advances the methodologies in OOD detection but also emphasizes the importance of dataset integrity for accurate algorithm evaluation.

cs.LG↗

arXiv2024

Roadmap for Unconventional Computing with Nanotechnology

In the "Beyond Moore's Law" era, with increasing edge intelligence, domain-specific computing embracing unconventional approaches will become increasingly prevalent. At the same time, adopting a variety of nanotechnologies will offer benefits in energy cost, computational speed, reduced footprint, cyber resilience, and processing power. The time is ripe for a roadmap for unconventional computing with nanotechnologies to guide future research, and this collection aims to fill that need. The authors provide a comprehensive roadmap for neuromorphic computing using electron spins, memristive devices, two-dimensional nanomaterials, nanomagnets, and various dynamical systems. They also address other paradigms such as Ising machines, Bayesian inference engines, probabilistic computing with p-bits, processing in memory, quantum memories and algorithms, computing with skyrmions and spin waves, and brain-inspired computing for incremental learning and problem-solving in severely resource-constrained environments. These approaches have advantages over traditional Boolean computing based on von Neumann architecture. As the computational requirements for artificial intelligence grow 50 times faster than Moore's Law for electronics, more unconventional approaches to computing and signal processing will appear on the horizon, and this roadmap will help identify future needs and challenges. In a very fertile field, experts in the field aim to present some of the dominant and most promising technologies for unconventional computing that will be around for some time to come. Within a holistic approach, the goal is to provide pathways for solidifying the field and guiding future impactful discoveries.

cs.ET↗

arXiv2023

Enhancing Few-shot CLIP with Semantic-Aware Fine-Tuning

Learning generalized representations from limited training samples is crucial for applying deep neural networks in low-resource scenarios. Recently, methods based on Contrastive Language-Image Pre-training (CLIP) have exhibited promising performance in few-shot adaptation tasks. To avoid catastrophic forgetting and overfitting caused by few-shot fine-tuning, existing works usually freeze the parameters of CLIP pre-trained on large-scale datasets, overlooking the possibility that some parameters might not be suitable for downstream tasks. To this end, we revisit CLIP's visual encoder with a specific focus on its distinctive attention pooling layer, which performs a spatial weighted-sum of the dense feature maps. Given that dense feature maps contain meaningful semantic information, and different semantics hold varying importance for diverse downstream tasks (such as prioritizing semantics like ears and eyes in pet classification tasks rather than side mirrors), using the same weighted-sum operation for dense features across different few-shot tasks might not be appropriate. Hence, we propose fine-tuning the parameters of the attention pooling layer during the training process to encourage the model to focus on task-specific semantics. In the inference process, we perform residual blending between the features pooled by the fine-tuned and the original attention pooling layers to incorporate both the few-shot knowledge and the pre-trained CLIP's prior knowledge. We term this method as Semantic-Aware FinE-tuning (SAFE). SAFE is effective in enhancing the conventional few-shot CLIP and is compatible with the existing adapter approach (termed SAFE-A).

cs.CV↗

arXiv2023

SR-OOD: Out-of-Distribution Detection via Sample Repairing

Out-of-distribution (OOD) detection is a crucial task for ensuring the reliability and robustness of machine learning models. Recent works have shown that generative models often assign high confidence scores to OOD samples, indicating that they fail to capture the semantic information of the data. To tackle this problem, we take advantage of sample repairing and propose a novel OOD detection framework, namely SR-OOD. Our framework leverages the idea that repairing an OOD sample can reveal its semantic inconsistency with the in-distribution data. Specifically, our framework consists of two components: a sample repairing module and a detection module. The sample repairing module applies erosion to an input sample and uses a generative adversarial network to repair it. The detection module then determines whether the input sample is OOD using a distance metric. Our framework does not require any additional data or label information for detection, making it applicable to various scenarios. We conduct extensive experiments on three image datasets: CIFAR-10, CelebA, and Pokemon. The results demonstrate that our approach achieves superior performance over the state-of-the-art generative methods in OOD detection.

cs.LG↗