Decoding Process Research Articles

Effective Bi-decoding networks for rail-surface defect detection by knowledge distillation

No-service rail-surface defect detection is a crucial method for assessing the quality of railroad tracks. However, the low-contrast and dark-tone characteristics of track-surface textures pose challenges to current defect-monitoring techniques. Real-time and on-site online inspections are important to ensure safe railway operation; however, most complex models for no-service inspections are difficult to deploy on mobile devices. To address these challenges and overcome the detection difficulties associated with complex scenes, we designed a knowledge distillation-based double decoding-layer refinement network (EBDNet-KD). The first decoding process is guided by a bimodal high-level semantic feature map obtained by extending the attention-based graph convolution to incrementally enhance the dual-stream features and obtain an image restoration prior. A divide-and-conquer decoder is then designed to distinguish features using different decoding layers. The prior is then used in the second decoding layer, which enables the bimodal features to interact fully and obtain the final prediction map. We introduce a knowledge distillation strategy that enables a lightweight, compact student network to learn a complex teacher network’s feature extraction process. This facilitates pixel-consistent learning of the knowledge within the bi-decoder layer, as well as bidirectional learning of the focused contextual response knowledge to optimize the model. The EBDNet-KD significantly reduces computational costs while guaranteeing performance with a parameter count of only 28 M. EBDNet-KD demonstrated superior performance over 15 state-of-the-art methods in experiments conducted on NEU RSDDS-AUG, an industrial RGB-depth dataset. We assessed the generalizability of EBDNet-KD by evaluating its performance on three additional public datasets, yielding competitive results. The source code and results can be found at https://github.com/Wuyue15/EBDNet.

Design and Signal-Decoding Test Verification of Dual-Channel Round Inductosyn Decoding Circuit

During the in-orbit operation of spacecraft, permanent magnet synchronous motors are commonly used as power sources in the drive mechanisms of solar panel arrays and the high-precision servo control systems based on satellites. Apart from the performance of the motors themselves and the software control algorithms, the accuracy of the entire control system is also influenced by angle sensors used to detect the rotor position of the motors. As a high-precision angular measuring instrument, the inductosyn possesses excellent environmental adaptability and long service life. Effectively utilizing the inductosyn can greatly enhance the performance of servo control systems. To address the complexity of the decoding process for dual-channel round inductosyn-to-digital converters, this paper proposes a design of the decoding circuit for dual-channel round inductosyn based on the parallel-synchronization decoding method of two AD2S1210 Resolver-to-Digital Converter (RDC) decoding chips. The decoding circuit amplifies the excitation signal outputted by the AD2S1210 for driving the round inductosyn, and processes the sine and cosine induction signals outputted by the round inductosyn through filtering, amplification, and other methods; by using analog circuitry, the output signals of the dual-channel round inductosyn are processed to meet the input requirements of the AD2S1210. Finally, through both the Multisim (circuit simulation software Version 14.1) simulation and physical experiments, it was verified that the decoding circuit designed in this paper could process the input/output signals of the dual-channel round inductosyn and AD2S1210, and successfully decoded the analog induction signal of the round inductosyn. This greatly simplifies the signal decoding process for the dual-channel round inductosyn.

Two-Phase Globally Coupled Low-Density Parity Check Decoding Aided with Early Termination and Forced Convergence.

To enhance the decoding efficiency of Globally Coupled (GC) LDPC codes, we incorporated Early Termination (ET) and Forced Convergence (FC) into the local/global two-phase decoding algorithm to expedite the decoding process. The two-phase decoding scheme integrates the ET technique to halt unnecessary iterations in the local decoding phase while employing the FC technique to accelerate convergence in the global phase decoding. The application of ET technology in the local decoding of GC-LDPC codes will not cause performance loss as in traditional block codes and will cause considerable complexity gains. For a longer code length and larger convergence differences between nodes' global codes, the FC technique operates more efficiently in global code than local code. Two variants are proposed for the ET scheme in the local decoding, namely ET-1 and ET-2. The initial variant, ET-1, predicts whether local decoding can be successful according to data characteristics and stop the local decoding iteration that is not expected to be successful in time. In the case of ET-2, the saved local iterations are transformed to global decoding equally. The results show that ET-1 saves considerable decoding time complexity and ET-2 improves the performance of the GC-LDPC code with the same decoding time complexity. The combined approach of ET-1 with FC reduces the decoding time complexity up to 42% at a low Signal Noise Rate region while maintaining its performance; ET-2-FC two-phase decoding saves approximately 25% decoding time complexity while improving the BER by about 0.18 dB and FER by about 0.23 dB.

Attention Induced Dual Convolutional-Capsule Network (AIDC-CN): A deep learning framework for motor imagery classification

In recent times, Electroencephalography (EEG)-based motor imagery (MI) decoding has garnered significant attention due to its extensive applicability in healthcare, including areas such as assistive robotics and rehabilitation engineering. Nevertheless, the decoding of EEG signals presents considerable challenges owing to their inherent complexity, non-stationary characteristics, and low signal-to-noise ratio. Notably, deep learning-based classifiers have emerged as a prominent focus for addressing the EEG signal decoding process. This study introduces a novel deep learning classifier named the Attention Induced Dual Convolutional-Capsule Network (AIDC-CN) with the specific aim of accurately categorizing various motor imagination class labels. To enhance the classifier’s performance, a dual feature extraction approach leveraging spectrogram and brain connectivity networks has been employed, diversifying the feature set in the classification task. The main highlights of the proposed AIDC-CN classifier includes the introduction of a dual convolution layer to handle the brain connectivity and spectrogram features, addition of a novel self-attention module (SAM) to accentuate the relevant parts of the convolved spectrogram features, introduction of a new cross-attention module (CAM) to refine the outputs obtained from the dual convolution layers and incorporation of a Gaussian Error Linear Unit (GELU) based dynamic routing algorithm to strengthen the coupling among the primary and secondary capsule layers. Performance analysis undertaken on four public data sets depict the superior performance of the proposed model with respect to the state-of-the-art techniques. The code for this model is available at https://github.com/RiteshSurChowdhury/AIDC-CN.

Contour-constrained branch U-Net for accurate left ventricular segmentation in echocardiography.

Using echocardiography to assess the left ventricular function is one of the most crucial cardiac examinations in clinical diagnosis, and LV segmentation plays a particularly vital role in medical image processing as many important clinical diagnostic parameters are derived from the segmentation results, such as ejection function. However, echocardiography typically has a lower resolution and contains a significant amount of noise and motion artifacts, making it a challenge to accurate segmentation, especially in the region of the cardiac chamber boundary, which significantly restricts the accurate calculation of subsequent clinical parameters. In this paper, our goal is to achieve accurate LV segmentation through a simplified approach by introducing a branch sub-network into the decoder of the traditional U-Net. Specifically, we employed the LV contour features to supervise the branch decoding process and used a cross attention module to facilitate the interaction relationship between the branch and the original decoding process, thereby improving the segmentation performance in the region LV boundaries. In the experiments, the proposed branch U-Net (BU-Net) demonstrated superior performance on CAMUS and EchoNet-dynamic public echocardiography segmentation datasets in comparison to state-of-the-art segmentation models, without the need for complex residual connections or transformer-based architectures. Our codes are publicly available at Anonymous Github https://anonymous.4open.science/r/Anoymous_two-BFF2/ .

Reducing the error rate of a superconducting logical qubit using analog readout information

Quantum error correction enables the preservation of logical qubits with a lower logical error rate than the physical error rate, with performance depending on the decoding method. Traditional decoding approaches rely on the binarization (“hardening”) of readout data, thereby ignoring valuable information embedded in the analog (“soft”) readout signal. We present experimental results showcasing the advantages of incorporating soft information into the decoding process of a distance-3 (d=3) bit-flip surface code with flux-tunable transmons. We encode each of the 16 computational states that make up the logical state |0L⟩, and protect them against bit-flip errors by performing repeated Z-basis stabilizer measurements. To infer the logical fidelity for the |0L⟩ state, we average across the 16 computational states and employ two decoding strategies: minimum-weight perfect matching and a recurrent neural network. Our results show a reduction of up to 6.8% in the extracted logical error rate with the use of soft information. Decoding with soft information is widely applicable, independent of the physical qubit platform, and could allow for shorter readout durations, further minimizing logical error rates. Published by the American Physical Society 2024

A semantic segmentation method integrated convolutional nonlinear spiking neural model with Transformer

Semantic segmentation is a critical task in computer vision, with significant applications in areas like autonomous driving and medical imaging. Transformer-based methods have gained considerable attention recently because of their strength in capturing global information. However, these methods often sacrifice detailed information due to the lack of mechanisms for local interactions. Similarly, convolutional neural network (CNN) methods struggle to capture global context due to the inherent limitations of convolutional kernels. To overcome these challenges, this paper introduces a novel Transformer-based semantic segmentation method called NSNPFormer, which leverages the nonlinear spiking neural P (NSNP) system—a computational model inspired by the spiking mechanisms of biological neurons. The NSNPFormer employs an encoding–decoding structure with two convolutional NSNP components and a residual connection channel. The convolutional NSNP components facilitate nonlinear local feature extraction and block-level feature fusion. Meanwhile, the residual connection channel helps prevent the loss of feature information during the decoding process. Evaluations on the ADE20K and Pascal Context datasets show that NSNPFormer achieves mIoU scores of 53.7 and 58.06, respectively, highlighting its effectiveness in semantic segmentation tasks.

Application of a novel visual representation method for urban noise source monitoring

Increasing urban mobility has led to a rise in the number of vehicles, resulting in urban noise pollution, which can cause health problems. Conventional methods for noise control can be adopted to reduce noise near the sources. However, this is not cost-effective to reduce overall noise level in the urban area. Therefore, it is preferable to encourage drivers to voluntarily reduce noise by applying methods such as cracking down on road noise sources. To this end, a real-time urban noise source monitoring system is needed, but processing large amounts of data can be difficult. In this study, a novel visual representation method, DoAgram, is applied for urban noise source monitoring. DoAgram can represent the time, frequency, and spatial information of the sound source within a 2D color graph. Therefore, the long-term spatial information of the sound source acquired from the sensor array node can be transmitted as a single image. Also, one can restore spatial information of the source from an image through a decoding process corresponding to the encoding method. One can find that the proposed method can be utilized as a database that can efficiently conduct urban noise source monitoring.

Implicit expression recognition enhanced table-filling for aspect sentiment triplet extraction

Aspect sentiment triplet extraction (ASTE) is a challenging task in aspect-based sentiment analysis (ABSA), involving the identification of aspect terms, opinion terms, and their corresponding sentiment polarities within comments to form triplets. The emergence of more realistic DMASTE datasets, featuring diverse domains, implicit aspect terms, and longer comments, poses challenges for existing methods. In particular, these methods struggle with recognizing implicit expressions effectively and capturing sufficient information. To overcome these hurdles, we propose an implicit expression recognition enhanced table-filling (IERET) method. This approach integrates modeling of overall implicit expression in sentences and employs a bidirectional information aggregation module to capture word pair information comprehensively. During the decoding process, a table-filling method accurately delineates aspect-opinion pair boundaries. Experimental results across in-domain, single-source cross-domain, and multi-source cross-domain on the DMASTE dataset demonstrate that our proposed IERET method achieves state-of-the-art performance.

Feature differences reduction and specific features preserving network for RGB-T salient object detection

In RGB-T salient object detection, effective utilization of the different characteristics of RGB and thermal modalities is essential to achieve accurate detection. Most of the previous methods usually only focus on reducing the differences between modalities, which may ignore the specific features that are crucial for salient object detection, leading to suboptimal results. To address the above issue, an RGB-T SOD network that simultaneously considers the reduction of modality differences and the preservation of specific features is proposed. Specifically, we construct a modality differences reduction and specific features preserving module (MDRSFPM) which aims to bridge the gap between modalities and enhance the specific features of each modality. In MDRSFPM, the dynamic vector generated by the interaction of RGB and thermal features is used to reduce modality differences, and then a dual branch is constructed to deal with the RGB and thermal modalities separately, employing a combination of channel-level and spatial-level operations to preserve their respective specific features. In addition, a multi-scale global feature enhancement module (MGFEM) is proposed to enhance global contextual information to provide guidance information for the subsequent decoding stage, so that the model can more easily localize the salient objects. Furthermore, our approach includes a fully fusion and gate module (FFGM) that utilises dynamically generated importance maps to selectively filter and fuse features during the decoding process. Extensive experiments demonstrate that our proposed model surpasses other state-of-the-art models on three publicly available RGB-T datasets remarkably. Our code will be released at https://github.com/JOOOOKII/FRPNet.

Retrieval In Decoder benefits generative models for explainable complex question answering

Large-scale Language Models (LLMs) utilizing the Chain-of-Thought prompting demonstrate exceptional performance in a variety of tasks. However, the persistence of factual hallucinations remains a significant challenge in practical applications. Prevailing retrieval-augmented methods treat the retriever and generator as separate components, which inadvertently restricts the generator’s capabilities to those of the retriever through intensive supervised training. In this work, we propose an unsupervised Retrieval In Decoder framework for multi-granularity decoding called RID, which integrates retrieval directly into the decoding process of generative models. It dynamically adjusts decoding granularity based on retrieval outcomes, and duly corrects the decoding direction through its direct impact on the next token. Moreover, we introduce a reinforcement learning-driven knowledge distillation method for adaptive explanation generation to better apply to Small-scale Language Models (SLMs). The experimental results across six public benchmarks surpass popular LLMs and existing retrieval-augmented methods, which demonstrates the effectiveness of RID in models of different scales and verifies its applicability and scalability.

Introducing bidirectional attention for autoregressive models in abstractive summarization

Abstractive summarization methods typically follow the autoregressive paradigm using the causal masks in the decoder for training and inference efficiency. However, this approach leads to a constant context throughout the generation process, which conflicting with the bidirectional characteristics of natural language. Although previous attempts have been made to incorporate bidirectional attention in the decoding process through non-autoregressive approach, the evaluation results are not comparable to the autoregressive methods. To bring bidirectional attention to the autoregressive process while maintaining superior performance, we propose the global autoregressive paradigm, which takes the outputs of the autoregressive process as additional inputs in the subsequent global iteration. Specifically, we build a bidirectional decoder alongside the original encoder and decoder to capture the bidirectional context of the outputs. This context is updated after each autoregressive decoding iteration. The decoder then integrates the updated context into subsequent autoregressive decoding steps, enhancing the generative process with a more comprehensive and authentic context. Additionally, we use contrastive learning to train the model to extract reliable features from the bidirectional context and apply reinforcement learning to improve the model's utilization of this context. We evaluate our method on CNN/DM, XSum, and NYT datasets, and the results highlight the significance of the bidirectional context. Our method achieves the best performance in terms of ROUGE-2 on CNN/DM (23.96), and performs comparably on XSum (25.45) and NYT (27.91). It also outperforms all the baselines in terms of BERTScore, with a score of 89.96 on CNN/DM, 92.70 on XSum, and 90.04 on NYT. Furthermore, our method can perform better with a larger beam size.

DBCAN: DFormer-Based Cross-Attention Network for RGB Depth Semantic Segmentation

Existing RGB-depth semantic segmentation methods primarily rely on symmetric two-stream Convolutional Neural Networks (CNNs) to extract RGB and spatial features separately. However, these architectures have limitations in incorporating spatial features and efficiently fusing RGB and depth information. In this study, we propose a novel architecture called the DFormer-Based Cross-Attention Network (DBCAN), which utilizes DFormer as an encoder for feature extraction and integrates several modifications to address these challenges. While DFormer is leveraged for its strong feature extraction capabilities, our modifications in the decoder focus on improving cross-modal fusion and spatial feature incorporation. We introduce three modules in the decoding process: the Object-Region Generated Module (ORGM), the Feature-Region Relation Module (FRRM), and the Spatial-Semantic Fusion Module (SSFM), which enhance feature interaction and segmentation accuracy. Experimental results on the NYUDepthv2 and SUN-RGBD datasets demonstrate that DBCAN achieves state-of-the-art performance, highlighting the effectiveness of our architectural enhancements in overcoming the limitations of existing models.

Embedded U-shaped network with cross-hierarchical feature adaptation fusion for remote sensing image haze removal

ABSTRACT Recently, U-Net architecture has been extensively explored for remote sensing (RS) image haze removal, demonstrating remarkable performance. However, most existing RS image haze removal methods based on the U-Net fail to fully utilize feature information extracted in the encoding layers during decoding, resulting in unsatisfactory haze removal results. Moreover, most haze removal algorithms neglect the interaction of cross-level feature information, which is particularly important for scene recovery and context information constraints. To address these issues, this paper presents an Embedded U-shaped Network with Cross-Hierarchical Feature Adaptation Fusion Network (EUCHA). Specifically, an Embedded U-shaped Network Framework (EUNF) is proposed to fuse the features extracted from different scale encoding layers as supplementary information into the corresponding decoding layers by embedding multiple parallel U-shaped subnetworks, which alleviates feature information dilution during the decoding process and enhances the network receptive field. Furthermore, we propose a Cross-Hierarchical Feature Adaptive Fusion (CHAF) module, which effectively and adaptively fuses multiple adjacent hierarchical features extracted by channel attention and pixel attention through learnable factors to enhance the expressive ability of features. Experiments were conducted on both real and synthetic datasets and compared with the nine most advanced algorithms. The results showed that EUCHA algorithm performs better in removing artefacts in both RS images and natural images. The source code is available at https://github.com/GongHongY/EUCHA.

Multi-GlaucNet: A multi-task model for optic disc segmentation, blood vessel segmentation and glaucoma detection

Glaucoma is a common and severe ocular disease that often leads to vision loss. The information on the optic disc (OD) and blood vessels in fundus images can significantly aid in glaucoma detection. In addition, the use of deep learning models for glaucoma detection is a highly effective approach. We propose a multi-task deep learning model called Multi-GlaucNet that can simultaneously segment the OD and blood vessels, thereby assisting doctors in diagnosing glaucoma. Multi-GlaucNet consists of three modules: the OD segmentation module, blood vessel segmentation module and glaucoma detection module. The OD segmentation module and blood vessel segmentation module both adopt an encoder–decoder structure to segment OD and blood vessels, respectively. The segmentation module is constructed using bottleneck layers in the encoding process and uses Pixel Shuffle and channel attention mechanisms in the decoding process. The detection module uses the ResNet50 network to perform glaucoma detection based on the features extracted from the segmentation modules. Multi-GlaucNet demonstrates outstanding performance in three areas. It achieves a Dice coefficient of 96.7% for OD segmentation on the ORIGA dataset, accuracy of 0.9798 and F1 score of 0.8562 for blood vessel segmentation on a mixed dataset. For glaucoma detection on the REFUGE dataset, it attains the highest accuracy of 0.967 and an area under the curve (AUC) of 0.950. These results validate the effectiveness of Multi-GlaucNet for glaucoma detection. The model’s ability to perform multiple tasks with high accuracy and efficiency demonstrates that it is a valuable aiding tool for glaucoma diagnosis.

Chaos-driven seven-core optical transmission scheme based on DNA full information chained analog-transcription

This paper proposes a chaos-driven seven-core optical transmission scheme based on DNA full information chained analog-transcription. Unlike traditional deoxyribonucleic acid (DNA) coded encryption schemes in the bit dimension, this scheme uses chaotic sequences to generate perturbed bit streams corresponding to the initial bit stream. These two sets of bit streams are encoded from a set of DNA double-stranded sequences, which are then intertwined into a single-stranded DNA containing all the information through the full-information class transcription algorithm proposed in this paper. Finally, the DNA decoding process is driven by a set of sequences derived from another chaotic model to transform the DNA sequence containing all information back into a bit sequence for subsequent transmission. Additional chaotic sequences interfere with the subcarriers, symbols, and constellation angles. Moreover, to maintain spectral efficiency, hiding the key in the frame header allows for the dynamic simultaneous transmission of signal and key. The transmission of encrypted 16 quadrature amplitude modulation-orthogonal chirp division multiplexing (16QAM-OCDM) signals is experimentally demonstrated at a net bit rate of 51.72 Gb/s over 2 km weakly coupled seven-core fiber. At the receiving end, the correct key decoder is able to accurately recover the data, while the bit error ratio (BER) at the illegal receiving end is 0.5. Finally, quantitative experiments validate the receiver-side decryption algorithm, showing that the proposed encryption scheme achieves a large key space of 10397. The key can be fully decoded when the optical power is above -20dBm. This scheme significantly enhances the security and flexibility of the communication system, making it a promising candidate for future optical communication physical layer encryption.

NPSFF-Net: Enhanced Building Segmentation in Remote Sensing Images via Novel Pseudo-Siamese Feature Fusion

Building segmentation has extensive research value and application prospects in high-resolution remote sensing image (HRSI) processing. However, complex architectural contexts, varied building morphologies, and non-building occlusions make building segmentation challenging. Compared with traditional methods, deep learning-based methods present certain advantages in terms of accuracy and intelligence. At present, the most popular option is to first apply a single neural network to encode an HRSI, then perform a decoding process through up-sampling or using a transposed convolution operation, and then finally obtain the segmented building image with the help of a loss function. Although effective, this approach not only tends to lead to a loss of detail information, but also fails to fully utilize the contextual features. As an alternative, we propose a novel network called NPSFF-Net. First, using an improved pseudo-Siamese network composed of ResNet-34 and ResNet-50, two sets of deep semantic features of buildings are extracted with the support of transfer learning, and four encoded features at different scales are obtained after fusion. Then, information from the deepest encoded feature is enriched using a feature enhancement module, and the resolutions are recovered via the operations of skip connections and transposed convolutions. Finally, the discriminative features of buildings are obtained using the designed feature fusion algorithm, and the optimal segmentation model is obtained by fitting a cross-entropy loss function. Our method obtained intersection-over-union values of 89.45% for the Aerial Imagery Dataset, 71.88% for the Massachusetts Buildings Dataset, and 68.72% for the Satellite Dataset I.

A path splitting and pruning strategy on list decoder for PAC codes

AbstractRecently, Arıkan proposed a polarization‐adjusted convolutional (PAC) codes, demonstrating their superior error correction performance over polar codes at short block lengths. It was confirmed that PAC codes approached the optimal performance achievable with limited code length. This paper proposes a novel low‐complexity list decoding algorithm for PAC codes, incorporating path splitting and pruning strategies based on a set of highly reliable information bits. Simulation results reveal that the proposed algorithm significantly reduces sorting complexity and average list size, all while incurring negligible performance loss. Unlike previous pruning algorithms designed for polar codes, the proposed strategy eliminates the need to individually assess the reliability of decoding paths in each decoding process. Instead, the algorithm minimizes redundant decoding paths through a high‐reliability information bit set, constructed using Monte Carlo experiments.

Image encryption algorithm based on multiple chaotic systems and improved Joseph block scrambling

Abstract With the rapid development of digital information technology, images are increasingly used in various fields. To ensure the security of image data, prevent unauthorized tampering and leakage, maintain personal privacy, and protect intellectual property rights, this study proposes an innovative color image encryption algorithm. Initially, the Mersenne Twister algorithm is utilized to generate high-quality pseudo-random numbers, establishing a robust basis for subsequent operations. Subsequently, two distinct chaotic systems, the autonomous non-Hamiltonian chaotic system and the tent-logistic-cosine chaotic mapping, are employed to produce chaotic random sequences. These chaotic sequences are used to control the encoding and decoding process of the DNA, effectively scrambling the image pixels. Furthermore, the complexity of the encryption process is enhanced through improved Joseph block scrambling. Thorough experimental verification, research, and analysis, the average value of the information entropy test data reaches as high as 7.999. Additionally, the average value of the number of pixels change rate (NPCR) test data is 99.6101%, which closely approaches the ideal value of 99.6094%. This algorithm not only guarantees image quality but also substantially raises the difficulty of decryption.

TabSAL: Synthesizing Tabular data with Small agent Assisted Language models

Tabular data are widely used in machine-learning tasks because of their prevalence in various fields; however, the potential risks of data breaches in tabular data and privacy protection regulations render such data almost unavailable. Tabular data generation methods alleviate data unavailability by synthesizing privacy-free data, and generating data using language models is a novel innovation. Language models can synthesize high-quality datasets by learning knowledge from nondestructive information and recognizing the semantics of table columns. However, when current language models function as generators, their encoding methods are hindered by complicated decoding processes, and the limited predictive ability of language models restricts their generative capability. To this end, we propose an encoding method based on interactive data structures such as JavaScript Object Notation for converting tabular data. We design TabSAL, which is a pluggable tabular data generation framework with small agent assisted language models, to boost the predictive capability, resulting in high-quality synthetic datasets with a much lower computational resource cost. In addition, a benchmark that integrates eight datasets, three methods, and three assessment directions has been issued, which indicates that TabSAL surpasses the state of the art by up to 60%.