AWAn-Yeu Wu

An-Yeu Wu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


An-Yeu Wu has a total of 94 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • PCG-Based Expectation Propagation Detector With Message-Passing Initialization and Approximated Variance for Fractional Doppler AFDM

    by Chi-Wei Chen, Yi-Ming Lee, Shao-Chun Wang, Chun-Hsuan Kuo, An-Yeu Wu

    Affine frequency division multiplexing (AFDM) has emerged as a promising Doppler-resilient waveform for high-mobility communications. To leverage its full diversity, we develop a real-valued expectation propagation (rEP) detector. However, its complexity is prohibitive, particularly in fractional Doppler scenarios with power leakages. Hence, we propose a Preconditioned Conjugate Gradient-based Message Passing-Initialized Approximated-Variance rEP (PCG-MPI-AVrEP), which employs diagonal reciprocity and single-pass MP for variance approximation. Based on channel structures, an incomplete LDL preconditioner is designed to accelerate convergence during PCG-based exact mean updates, reducing iterations by at least 86% compared to CG. Simulation results demonstrate that PCG-MPI-AVrEP significantly outperforms prior detectors while maintaining quadratic complexity with respect to subcarriers.


  • Open Access

    Quick Bypass Mechanism of Zero-Shot Diffusion-Based Image Restoration

    by Yu-Shan Tai, An- Yeu, Andy Wu

    Recent advancements in diffusion models have demonstrated remarkable success in various image generation tasks. Building upon these achievements, diffusion models have also been effectively adapted to image restoration tasks, e.g., super-resolution and deblurring, aiming to recover high-quality images from degraded inputs. Although existing zero-shot approaches enable pretrained diffusion models to perform restoration tasks without additional fine-tuning, these methods often suffer from prolonged iteration times in the denoising process. To address this limitation, we propose a Quick Bypass Mechanism (QBM), a strategy that significantly accelerates the denoising process by initializing from an intermediate approximation, effectively bypassing early denoising steps. Furthermore, recognizing that approximation may introduce inconsistencies, we introduce a Revised Reverse Process (RRP), which adjusts the weighting of random noise to enhance the stochasticity and mitigate potential disharmony. We validate proposed methods on ImageNet-1K and CelebA-HQ across multiple image restoration tasks, e.g., super-resolution, deblurring, and compressed sensing. Our experimental results show that the proposed methods can effectively accelerate existing methods while maintaining original performance.

  • Open Access

    Efficient and Reliable Vector Similarity Search Using Asymmetric Encoding with NAND-Flash for Many-Class Few-Shot Learning

    by Hao-Wei Chiang, Chi-Tse Huang, Hsiang-Yun Cheng, Po-Hao Tseng, Ming-Hsiu Lee, An-Yeu (Andy) Wu

    While memory-augmented neural networks (MANNs) offer an effective solution for few-shot learning (FSL) by integrating deep neural networks with external memory, the capacity requirements and energy overhead of data movement become enormous due to the large number of support vectors in many-class FSL scenarios. Various in-memory search solutions have emerged to improve the energy efficiency of MANNs. NAND-based multi-bit content addressable memory (MCAM) is a promising option due to its high density and large capacity. Despite its potential, MCAM faces limitations such as a restricted number of word lines, limited quantization levels, and non-ideal effects like varying string currents and bottleneck effects, which lead to significant accuracy drops. To address these issues, we propose several innovative methods. First, the Multi-bit Thermometer Code (MTMC) leverages the extensive capacity of MCAM to enhance vector precision using cumulative encoding rules, thereby mitigating the bottleneck effect. Second, the Asymmetric Vector Similarity Search (AVSS) reduces the precision of the query vector while maintaining that of the support vectors, thereby minimizing the search iterations and improving efficiency in many-class scenarios. Finally, the Hardware-Aware Training (HAT) method optimizes controller training by modeling the hardware characteristics of MCAM, thus enhancing the reliability of the system. Our integrated framework reduces search iterations by up to 32×, and increases overall accuracy by 1.58% to 6.94%.


  • Relay-Assisted Carrier Aggregation (RACA) Uplink System for Enhancing Data Rate of Extended Reality (XR)

    by Chi-Wei Chen, Wen-Chiao Tsai, Lung-Sheng Tsai, An-Yeu Wu

    In Extended Reality (XR) applications, high data rates and low latency are crucial for immersive experiences. Uplink transmission in XR is challenging due to the limited antennas and power of lightweight XR devices. To improve data transmission rates, we investigate a relay-assisted carrier aggregation (RACA) system. The XR device simultaneously transmits data to an access point (AP) and a relay in proximity over low-frequency and high-frequency bands, respectively. Then, the relay down-converts and amplifies the signals to the AP, effectively acting as an additional transmit antenna for the XR device. In this paper, we propose two algorithms to maximize the data rate of the XR device in their respective protocols. In the centralized protocol, the rate maximization problem is equivalently transformed as a weighted mean square error minimization (WMMSE) problem which can be solved iteratively by alternative optimization. In the distributed protocol, the rate maximization problem is decomposed into two independent subproblems where the rate of the direct link and the rate of the relay link are maximized by singular value decomposition (SVD)- based methods with water-filling (WF). Simulation results show that the rate of the RACA system is improved by 32% compared to that of the conventional carrier aggregation scheme.


  • EECS: Efficient End-Cloud Collaborative System With On-Demand Offloading Mechanism for Convolutional Neural Network

    by Yi-Cheng Lo, Cheng-Lin Hsieh, Hao-Wei Chiang, An-Yeu Wu

    The end-cloud synergy with neural networks offers significant promise for enhancing AI-IoT systems. Numerous studies have been conducted to effectively address the dynamic and resource-constrained nature of IoT; however, challenges persist in adaptability to constraints, horizontal integration between technologies, and system-level optimization. In this paper, we propose an end-cloud collaborative system (EECS) with an on-demand offloading mechanism for AI-IoT. We enhance three critical areas previously narrowly explored: model pruning to fit into the end device’s constraints, dynamic data transmission between end and cloud, and an effective offloading policy. We pay particular attention to the constraints imposed by multiply-and-add (MAC) operations and bandwidth (BW) at the end, with a focus on image classification tasks. Our contributions are threefold. First, we introduce a two-stage trainable pruning method that can automatically adjust the end model to end constraints and optimize from a system-wide perspective. Second, we propose an adaptive mechanism that accommodates fluctuating bandwidth conditions based on our trainable pruning, integrating static pruning with dynamic inference. Third, we develop a smart offloading policy that enhances decision-making, thereby elevating system-level efficiency. Finally, our EECS shows substantial improvements, achieving 2.6×~4× reductions in MAC costs and 3.8×~10.4× reductions in BW costs compared to existing approaches, based on evaluations using the CIFAR-100 and Tiny-ImageNet-200 datasets.


  • Open Access

    An Efficient Anomalous Sound Detection System for Microcontrollers

    by Yi-Cheng Lo, Tsung-Lin Tsai, Chieh-Wen Yang, An-Yeu Wu

    Anomalous Sound Detection (ASD) systems are pivotal in the Industrial Internet of Things (IIoT). Through the early detection of machines’ anomalies, these systems facilitate proactive maintenance, thereby mitigating potential losses. Although prior studies have improved system accuracy using various advanced machine learning technologies, they frequently neglect the associated substantial computing and storage demands, which are crucial in resource-constrained IIoT environments. In this paper, we propose an ASD system that is efficiently optimized for both software and hardware considerations regarding edge intelligence. For the software aspect, we identify signal variation as a critical issue for ASD. Hence, we introduce a suite of lightweight yet robust processing techniques that enhance accuracy while minimizing resource consumption. As for the hardware aspect, we find that memory constraints may be a significant challenge for deploying ASD systems on microcontrollers (MCUs). Therefore, we propose a memory-aware pruning algorithm specialized for ASD to fit into MCUs’ constraints. Finally, we evaluate our method on the DCASE dataset, and the results show that our system achieves favorable outcomes in both accuracy and resource efficiency, marking our contribution to ASD system practice.


  • Open Access

    Efficient and Reliable Vector Similarity Search Using Asymmetric Encoding with NAND-Flash for Many-Class Few-Shot Learning

    by Hao-Wei Chiang, Chi-Tse Huang, Hsiang-Yun Cheng, Po-Hao Tseng, Ming-Hsiu Lee, An- Yeu, Andy Wu

    While memory-augmented neural networks (MANNs) offer an effective solution for few-shot learning (FSL) by integrating deep neural networks with external memory, the capacity requirements and energy overhead of data movement become enormous due to the large number of support vectors in many-class FSL scenarios. Various in-memory search solutions have emerged to improve the energy efficiency of MANNs. NAND-based multi-bit content addressable memory (MCAM) is a promising option due to its high density and large capacity. Despite its potential, MCAM faces limitations such as a restricted number of word lines, limited quantization levels, and non-ideal effects like varying string currents and bottleneck effects, which lead to significant accuracy drops. To address these issues, we propose several innovative methods. First, the Multi-bit Thermometer Code (MTMC) leverages the extensive capacity of MCAM to enhance vector precision using cumulative encoding rules, thereby mitigating the bottleneck effect. Second, the Asymmetric vector similarity search (AVSS) reduces the precision of the query vector while maintaining that of the support vectors, thereby minimizing the search iterations and improving efficiency in many-class scenarios. Finally, the Hardware-Aware Training (HAT) method optimizes controller training by modeling the hardware characteristics of MCAM, thus enhancing the reliability of the system. Our integrated framework reduces search iterations by up to 32 times, and increases overall accuracy by 1.58% to 6.94%.

  • Deep Unfolding-Based Channel Estimation for IRS-Aided mmWave Systems via Two-Stage LAMP Network With Row Compression

    by Wen-Chiao Tsai, Chi-Wei Chen, An-Yeu Wu

    Intelligent reflecting surface (IRS) is emerging as a promising and revolutionary technology for achieving cost-effective wireless communication systems. However, the performance of IRS communications heavily relies on acquiring accurate channel state information (CSI), which is challenging under low training overhead due to a large number of passive IRS elements. This paper proposes a two-stage LAMP network with row compression (RCTS-LAMP) to solve the joint estimation problem of direct and cascaded channels in IRS-aided millimeter-wave (mmWave) systems. Specifically, the proposed RCTS-LAMP is a model-driven neural network that combines the advantages of compressive sensing (CS) and deep learning (DL) by using the deep unfolding technique. By doing so, we can recover the direct and cascaded channels with CS under low training overhead, and the estimation performance can be significantly improved with the joint optimization of DL. Meanwhile, the cascaded channel estimation is decomposed into two stages to reduce computational complexity further. Numerical results show that the RCTS-LAMP network can estimate the cascaded channel with a better trade-off between computational complexity and accuracy, while the direct channel can be jointly recovered without adding an extra network.


  • Open Access

    Quality-Aware Signal Processing Mechanism of PPG Signal for Long-Term Heart Rate Monitoring

    by Win-Ken Beh, Yu-Chia Yang, An-Yeu Wu

    Photoplethysmography (PPG) is widely utilized in wearable healthcare devices due to its convenient measurement capabilities. However, the unrestricted behavior of users often introduces artifacts into the PPG signal. As a result, signal processing and quality assessment play a crucial role in ensuring that the information contained in the signal can be effectively acquired and analyzed. Traditionally, researchers have discussed signal quality and processing algorithms separately, with individual algorithms developed to address specific artifacts. In this paper, we propose a quality-aware signal processing mechanism that evaluates incoming PPG signals using the signal quality index (SQI) and selects the appropriate processing method based on the SQI. Unlike conventional processing approaches, our proposed mechanism recommends processing algorithms based on the quality of each signal, offering an alternative option for designing signal processing flows. Furthermore, our mechanism achieves a favorable trade-off between accuracy and energy consumption, which are the key considerations in long-term heart rate monitoring.