arXiv subjects

Hongzhi Zhu

Publications and source records attributed to Hongzhi Zhu.

6 recordsLinked to original sources

arXiv2026

A general performance analysis framework for bitwise neural polar decoding

This paper investigates when the population mean-squared error (MSE) of a bitwise neural predictor is sufficient to guarantee a bit-error rate (BER) close to the successive-cancellation (SC) reference performance. Each synthesized message channel is formulated as a soft maximum a posteriori (MAP) regression problem. Architecture-independent direct and posterior-margin MSE-to-BER bounds then convert certified population MSE into an SC-referenced reliability requirement, while an independent held-out certificate makes the population condition verifiable from data. A three-layer over-parameterized neural network (ONN) decoder, comprising an input layer, one over-parameterized rectified linear unit (ReLU) hidden layer, and a fixed-sign output layer, provides a constructive instance with provable empirical MSE convergence under explicit full-batch gradient-descent conditions. A (128,64) shared-message-update polar factor-graph neural decoder is further evaluated over additive white Gaussian noise (AWGN) and block-Rayleigh channels to examine architecture-independent certification and engineering performance. The resulting reliability-certification framework connects trainability, population regression accuracy, and communication reliability. It provides verifiable sufficient conditions for an SC-referenced BER guarantee and its associated block-error rate (BLER) characterization.

eess.SP↗

arXiv2025

Latent-attention Based Transformer for Near ML Polar Decoding in Short-code Regime

Transformer architectures have emerged as promising deep learning (DL) tools for modeling complex sequence-to-sequence interactions in channel decoding. However, current transformer-based decoders for error correction codes (ECCs) demonstrate inferior performance and generalization capabilities compared to conventional algebraic decoders, especially in short-code regimes. In this work, we propose a novel latent-attention based transformer (LAT) decoder for polar codes that addresses the limitations on performance and generalization through three pivotal innovations. First, we develop a latent-attention mechanism that supersedes the conventional self-attention mechanism. This architectural modification enables independent learning of the Query and Key matrices for code-aware attention computation, decoupling them from the Value matrix to emphasize position-wise decoding interactions while reducing context correlation interference. Second, we devise an advanced training framework incorporating three synergistic components: entropy-aware importance sampling that emphasizes low-probability regions in the signal constellation space, experience reflow that introduces empirical labels to improve characterization of decoding boundaries, and dynamic label smoothing for likelihood-based regularization. Third, we propose a code-aware mask scheme which allows dynamic adaptation for varying code configurations. Numerical evaluations demonstrate that the proposed LAT decoder achieves near maximum-likelihood (ML) performance in terms of both bit error rate (BER) and block error rate (BLER) for short-length polar codes. Furthermore, the architecture exhibits robust generalization capabilities across diverse code rates and code lengths.

eess.SP↗

arXiv2023

Semi-supervised MIMO Detection Using Cycle-consistent Generative Adversarial Network

In this paper, a new semi-supervised deep multiple-input multiple-output (MIMO) detection approach using a cycle-consistent generative adversarial network (CycleGAN) is proposed for communication systems without any prior knowledge of underlying channel distributions. Specifically, we propose the CycleGAN detector by constructing a bidirectional loop of two modified least squares generative adversarial networks (LS-GAN). The forward LS-GAN learns to model the transmission process, while the backward LS-GAN learns to detect the received signals. By optimizing the cycle-consistency of the transmitted and received signals through this loop, the proposed method is trained online and semi-supervisedly using both the pilots and the received payload data. As such, the demand on labelled training dataset is considerably controlled, and thus the overhead is effectively reduced. Numerical results show that the proposed CycleGAN detector achieves better performance in terms of both bit error-rate (BER) and achievable rate than existing semi-blind deep learning (DL) detection methods as well as conventional linear detectors, especially when considering signal distortion due to the nonlinearity of power amplifiers (PA) at the transmitter.

eess.SP↗

arXiv2022

ResNet Structure Simplification with the Convolutional Kernel Redundancy Measure

Deep learning, especially convolutional neural networks, has triggered accelerated advancements in computer vision, bringing changes into our daily practice. Furthermore, the standardized deep learning modules (also known as backbone networks), i.e., ResNet and EfficientNet, have enabled efficient and rapid development of new computer vision solutions. Yet, deep learning methods still suffer from several drawbacks. One of the most concerning problems is the high memory and computational cost, such that dedicated computing units, typically GPUs, have to be used for training and development. Therefore, in this paper, we propose a quantifiable evaluation method, the convolutional kernel redundancy measure, which is based on perceived image differences, for guiding the network structure simplification. When applying our method to the chest X-ray image classification problem with ResNet, our method can maintain the performance of the network and reduce the number of parameters from over $23$ million to approximately $128$ thousand (reducing $99.46\%$ of the parameters).

cs.CV↗

arXiv2022

Gaze-Guided Class Activation Mapping: Leveraging Human Attention for Network Attention in Chest X-rays Classification

The increased availability and accuracy of eye-gaze tracking technology has sparked attention-related research in psychology, neuroscience, and, more recently, computer vision and artificial intelligence. The attention mechanism in artificial neural networks is known to improve learning tasks. However, no previous research has combined the network attention and human attention. This paper describes a gaze-guided class activation mapping (GG-CAM) method to directly regulate the formation of network attention based on expert radiologists' visual attention for the chest X-ray pathology classification problem, which remains challenging due to the complex and often nuanced differences among images. GG-CAM is a lightweight ($3$ additional trainable parameters for regulating the learning process) and generic extension that can be easily applied to most classification convolutional neural networks (CNN). GG-CAM-modified CNNs do not require human attention as an input when fully trained. Comparative experiments suggest that two standard CNNs with the GG-CAM extension achieve significantly greater classification performance. The median area under the curve (AUC) metrics for ResNet50 increases from $0.721$ to $0.776$. For EfficientNetv2 (s), the median AUC increases from $0.723$ to $0.801$. The GG-CAM also brings better interpretability of the network that facilitates the weakly-supervised pathology localization and analysis.

eess.IV↗

arXiv2022

Multi-task UNet: Jointly Boosting Saliency Prediction and Disease Classification on Chest X-ray Images

Human visual attention has recently shown its distinct capability in boosting machine learning models. However, studies that aim to facilitate medical tasks with human visual attention are still scarce. To support the use of visual attention, this paper describes a novel deep learning model for visual saliency prediction on chest X-ray (CXR) images. To cope with data deficiency, we exploit the multi-task learning method and tackles disease classification on CXR simultaneously. For a more robust training process, we propose a further optimized multi-task learning scheme to better handle model overfitting. Experiments show our proposed deep learning model with our new learning scheme can outperform existing methods dedicated either for saliency prediction or image classification. The code used in this paper is available at https://github.com/hz-zhu/MT-UNet.

eess.IV↗