Classification Module Research Articles

Classification of peanut pod rot based on improved YOLOv5s.

Peanut pod rot is one of the major plant diseases affecting peanut production and quality over China, which causes large productivity losses and is challenging to control. To improve the disease resistance of peanuts, breeding is one significant strategy. Crucial preventative and management measures include grading peanut pod rot and screening high-contributed genes that are highly resistant to pod rot should be carried out. A machine vision-based grading approach for individual cases of peanut pod rot was proposed in this study, which avoids time-consuming, labor-intensive, and inaccurate manual categorization and provides dependable technical assistance for breeding studies and peanut pod rot resistance. The Shuffle Attention module has been added to the YOLOv5s (You Only Look Once version 5 small) feature extraction backbone network to overcome occlusion, overlap, and adhesions in complex backgrounds. Additionally, to reduce missing and false identification of peanut pods, the loss function CIoU (Complete Intersection over Union) was replaced with EIoU (Enhanced Intersection over Union). The recognition results can be further improved by introducing grade classification module, which can read the information from the identified RGB images and output data like numbers of non-rotted and rotten peanut pods, the rotten pod rate, and the pod rot grade. The Precision value of the improved YOLOv5s reached 93.8%, which was 7.8%, 8.4%, and 7.3% higher than YOLOv5s, YOLOv8n, and YOLOv8s, respectively; the mAP (mean Average Precision) value was 92.4%, which increased by 6.7%, 7.7%, and 6.5%, respectively. Improved YOLOv5s has an average improvement of 6.26% over YOLOv5s in terms of recognition accuracy: that was 95.7% for non-rotted peanut pods and 90.8% for rotten peanut pods. This article presented a machine vision- based grade classification method for peanut pod rot, which offered technological guidance for selecting high-quality cultivars with high resistance to pod rot in peanut.

Compound fault diagnosis of diesel engines by combining generative adversarial networks and transfer learning

In order to solve the problem of compound fault diagnosis of diesel engine fuel injection system under the condition of few samples, a comprehensive diagnosis method based on generative adversarial networks and transfer learning based multi label classification models is proposed in this paper, which is based on convolutional attention networks. The generative adversarial networks are used to enhance the original data, and the enhanced data is used as the source data for transfer learning to ensure its effectiveness. Design multi-dimensional labels for compound fault diagnosis to enable the model to diagnose compound faults and their individual faults. Pre train the convolutional attention network with enhanced data generated from adversarial networks. Subsequently, the original data is used to integrate the feature information in the network and perform secondary training on the classification module. Finally, a compound fault diagnosis is implemented based on the model after secondary training. The effectiveness of the comprehensive diagnostic method proposed in the article is verified using real fault data of the fuel injection system. The diagnostic accuracy on test data reached 80%. The effectiveness of the comprehensive diagnostic method proposed in the article was verified using real fault data of the fuel injection system. The diagnostic accuracy on test data reached 80%. Compared to the method of directly training convolutional attention networks using raw data and the method of directly training convolutional attention networks without transfer learning using GAN enhanced datasets, the accuracy is improved by 16% and 8%, respectively.

Improving Apple Fruit Quality Detection with AI and Machine Vision

The detection of apples using Raspberry Pi is an innovative approach that merges the realms of computer vision, machine learning, and agricultural automation. This abstract provides an extensive overview of the methodologies, implementations, challenges, and future directions pertaining to apple detection using Raspberry Pi, encapsulating the essence of the research conducted in this domain. The quest for automation in agriculture has spurred the development of novel technologies aimed at improving efficiency and reducing manual labor. Fruit detection, particularly the identification of apples, holds significant importance due to the fruit's widespread cultivation and economic value. Traditional methods of fruit detection often involve manual sorting, which is labor>intensive and time>consuming. Hence, there arises a need for automated systems capable of accurately identifying and sorting fruits, thereby streamlining agricultural processes. The implementation section details the practical realization of the apple detection system using Raspberry Pi. Hardware setup involves the integration of Raspberry Pi boards with camera modules and other peripherals necessary for image acquisition and processing. Software development entails the creation of Python>based modules for image preprocessing, feature extraction, and classification. OpenCV and scikit>learn libraries are utilized for implementing image processing and machine learning algorithms, respectively. The system is tested in different environments to evaluate its performance under various conditions, including controlled laboratory settings and outdoor agricultural scenarios.

Enhancing Insect Sound Classification Using Dual-Tower Network: A Fusion of Temporal and Spectral Feature Perception

In the modern field of biological pest control, especially in the realm of insect population monitoring, deep learning methods have made further advancements. However, due to the small size and elusive nature of insects, visual detection is often impractical. In this context, the recognition of insect sound features becomes crucial. In our study, we introduce a classification module called the “dual-frequency and spectral fusion module (DFSM)”, which enhances the performance of transfer learning models in audio classification tasks. Our approach combines the efficiency of EfficientNet with the hierarchical design of the Dual Towers, drawing inspiration from the way the insect neural system processes sound signals. This enables our model to effectively capture spectral features in insect sounds and form multiscale perceptions through inter-tower skip connections. Through detailed qualitative and quantitative evaluations, as well as comparisons with leading traditional insect sound recognition methods, we demonstrate the advantages of our approach in the field of insect sound classification. Our method achieves an accuracy of 80.26% on InsectSet32, surpassing existing state-of-the-art models by 3 percentage points. Additionally, we conducted generalization experiments using three classic audio datasets. The results indicate that DFSM exhibits strong robustness and wide applicability, with minimal performance variations even when handling different input features.

Tang Chang’an poetry automatic classification: a practical application of deep learning methods

Abstract As the capital of Tang Dynasty, Chang’an was one of the most prosperous cities in the world at that time and had a profound influence on Tang poetry. Poets described Chang’an to illustrate the cultural features of the Tang Dynasty while also invoking emotions in readers. The study of Tang Chang’an poetry has important literary and historical value. In order to understand the interpretation and emotional expression of Tang Chang’an poetry more conveniently and clearly, we conducted a study using deep learning to classify Chang’an poetry into four classes: imperially assigned poetry (应制), emotional poetry (感怀), parting poetry (离别), and other poetry (其他). We suggested a comprehensive framework of text classification based on deep learning, including a text input module, feature encoder module, and classification module. We applied several mainstream deep neural network structures to extract features in different ways, which comprised convolutional neural network (CNN), Fasttext, bi-direction long-short-term memory network, and Attention mechanism. Based on our experimental findings, the CNN-based method achieved the best performance for the task. Our inference was that, in Chinese ancient poetry, the analysis of semantic content is more facilitated by local textual features rather than contextual features. We combined this inference with the theory of image in Chinese ancient poetry to analyze the suitability of the deep learning techniques for the study of Chinese ancient poetry.

A two‐stage substation equipment classification method based on dual‐scale attention

AbstractAccurate classification of substation equipment images remains challenging due to various factors such as unexpected illumination, viewing angles, scale variations, shadows, surface contaminants, and different elements sharing similar appearances. This paper presents a novel two‐stage substation equipment classification method based on dual‐scale attention. Leveraging the region proposal technique from Faster‐regions with CNN features (RCNN), the input images are initially decomposed into multiple scales to capture latent features. A dual‐scale attention module is introduced to enhance the precision of feature extraction. Furthermore, a two‐stage network is proposed to address the challenge of classifying closely similar substation equipment. A multi‐layer perceptron performs a coarse classification to categorize the equipment into broad categories. Then, a lightweight classifier is employed for fine‐grained subclassification, further distinguishing equipment within the same broad category. To mitigate the issue of limited training data, a specialized dataset is collected and annotated for the substation equipment classification. Experimental results demonstrate that the proposed method achieves remarkable accuracy, recall, and F1‐score surpassing 0.91, outperforming mainstream approaches in terms of recall and F1 scores. Ablation experiments further validate the significant contributions of both the dual‐scale attention and the two‐stage classification module in improving the overall performance of the classification network.

Classification-driven intelligent system for automated evaluation of higher education exam paper quality

AbstractEvaluating the quality of university exam papers is crucial for universities seeking institutional and program accreditation. Currently, exam papers are assessed manually, a process that can be tedious, lengthy, and in some cases, inconsistent. This is often due to the focus on assessing only the formal specifications of exam papers. This study develops an intelligent system for the automatic evaluation of university exam papers in terms of form and content, ensuring adherence to quality standards. The system is composed of two subsystems: the first evaluates compliance with formal specifications, and the second analyzes the content. The content analysis involves automatically categorizing exam questions based on Bloom's cognitive levels (BCLs) and determining the representation ratio of these levels in the exam paper. This subsystem comprises four main modules: 1) question collection, 2) text pre-processing using natural language processing (NLP) methods, 3) feature engineering using the CountVectorizer method to convert questions into feature vectors, and 4) a classification module based on the Logistic Regression (LR) algorithm to categorize exam questions into categories like knowledge, comprehension, application, analysis, synthesis, and evaluation. Experimental results indicate that the system achieves an average accuracy of 98.5%.

Enhancing mass spectrometry data analysis: A novel framework for calibration, outlier detection, and classification

Mass spectrometry (MS) is a powerful analytical technique in metabolomics, enabling the identification and quantification of metabolites. However, analyzing MS data poses challenges such as batch effects, outliers, and high-dimensional data. In this paper, we propose a comprehensive framework for MS data analysis. The framework integrates data calibration, outlier detection, and automatic classification modules. Data calibration is performed using a deep autoencoder to remove batch effects. Outlier detection combines multiple algorithms through ensemble learning to identify and remove outliers. Automatic classification utilizes a transformer model to handle high-dimensional data and capture global feature relationships. Experimental results on myocardial infarction (MI) and coronary heart disease (CHD) datasets demonstrate the effectiveness of the framework. It outperforms traditional classification models and achieves higher accuracy. The proposed framework provides a robust solution for MS data analysis, facilitating more accurate classification and enabling reliable biological insights in metabolomics research.

Hand-held GPU accelerated device for multiclass classification of X-ray images using CNN model

Chest X-ray (CXR) images are the primary investigation aid for many lung diseases and their follow-ups. For diagnosis of SARS-CoV-2, RT–PCR test and chest Computed Tomography (CT) are commonly used but both face false negatives for ruling out the infection. So, there is a demanding need for developing a system combined with Artificial Intelligence (AI) and CXR imaging to detect COVID-19 patients to avoid its spread. Here, a robust and efficient handheld device is proposed. It uses the computational power of the Graphics Processing Unit (GPU) and pre-trained deep learning models for analyzing the CXR images. A Resnet-50 CNN model is deployed on an NVIDIA Jetson Nano GPU module for the real-time classification of COVID-19, Tuberculosis, and Normal using CXR images. The device can perform real-time classification of CXR images from a portable X-ray machine and classify the image into one of the above categories. For the extensive training, a database of 680 COVID-19, 1230 Tuberculosis, and 1050 normal CXR images are extracted by combining several global databases like Kaggle, SIRM, RSNA, and Radiopaedia. The classification accuracy, precision, and loss rate were 0.9879, 0.9758, and 0.0196 respectively and our model would improve with larger data sets. The highly accurate and high-performance GPU device significantly plays a far-reaching role in COVID-19 diagnosis using Chest X-ray, which could be beneficial to triage the health system and to combat the outbreak of COVID-19.

Learning to Synthesize Compatible Fashion Items Using Semantic Alignment and Collocation Classification: An Outfit Generation Framework.

The field of fashion compatibility learning has attracted great attention from both the academic and industrial communities in recent years. Many studies have been carried out for fashion compatibility prediction, collocated outfit recommendation, artificial intelligence (AI)-enabled compatible fashion design, and related topics. In particular, AI-enabled compatible fashion design can be used to synthesize compatible fashion items or outfits to improve the design experience for designers or the efficacy of recommendations for customers. However, previous generative models for collocated fashion synthesis have generally focused on the image-to-image translation between fashion items of upper and lower clothing. In this article, we propose a novel outfit generation framework, i.e., OutfitGAN, with the aim of synthesizing a set of complementary items to compose an entire outfit, given one extant fashion item and reference masks of target synthesized items. OutfitGAN includes a semantic alignment module (SAM), which is responsible for characterizing the mapping correspondence between the existing fashion items and the synthesized ones, to improve the quality of the synthesized images, and a collocation classification module (CCM), which is used to improve the compatibility of a synthesized outfit. To evaluate the performance of our proposed models, we built a large-scale dataset consisting of 20 000 fashion outfits. Extensive experimental results on this dataset show that our OutfitGAN can synthesize photo-realistic outfits and outperform the state-of-the-art methods in terms of similarity, authenticity, and compatibility measurements.

Enhancing temporal action localization in an end-to-end network through estimation error incorporation

Temporal action localization presents a significant challenge in computer vision, as the development of an efficient method for this task remains elusive. The objective is to identify human activities within untrimmed videos, determining when and which actions occur in each video. While using trimmed videos could potentially resolve the localization problem and enhance classification accuracy, it is impractical for real-world applications as the trimming process itself requires human intervention. This highlights the importance of temporal localization. Due to the availability of several successful approaches for action recognition in trimmed video, conventional multi-stage methods for untrimmed video, commonly employ a network to generate activity proposals, followed by a separate network for classification. These disjoint networks are optimized individually and thus usually vary from the global optimum, leading to less precise candidate action proposals. To address this challenge, we propose a novel end-to-end neural network that utilizes error estimation for precise action localization and recognition in untrimmed videos. The proposed method performs the localization and classification of action instances simultaneously, thereby optimizing the corresponding networks concurrently. To increase the precision of the action proposal boundaries, the Regression module is innovatively utilized as part of the proposed end-to-end network, along with the Evaluation and Classification modules. This module estimates the potential error in proposal time boundaries and enhances the result accuracy. We have conducted experiments on THUMOS 14 and ActivityNet-1.3, which are considered the most challenging datasets for temporal action localization. The novel, yet fairly simple, proposed network achieves remarkable performance improvement compared to the other state-of-the-art methods. This improvement, which is more pronounced in the cases of high temporal intersection with ground truth, is accomplished without requiring extra data or complicated architecture. By incorporating error estimation, we achieved improvement in mean Average Precision (mAP). The proposed approach particularly shines for the localization of challenging activities in the complex and diverse dataset ActivityNet-1.3. For instance, for the “drinking coffee” activity, the mean Average Precision (mAP) was enhanced fivefold compared to the best-reported results.

TransPTM: a transformer-based model for non-histone acetylation site prediction.

Protein acetylation is one of the extensively studied post-translational modifications (PTMs) due to its significant roles across a myriad of biological processes. Although many computational tools for acetylation site identification have been developed, there is a lack of benchmark dataset and bespoke predictors for non-histone acetylation site prediction. To address these problems, we have contributed to both dataset creation and predictor benchmark in this study. First, we construct a non-histone acetylation site benchmark dataset, namely NHAC, which includes 11 subsets according to the sequence length ranging from 11 to 61 amino acids. There are totally 886 positive samples and 4707 negative samples for each sequence length. Secondly, we propose TransPTM, a transformer-based neural network model for non-histone acetylation site predication. During the data representation phase, per-residue contextualized embeddings are extracted using ProtT5 (an existing pre-trained protein language model). This is followed by the implementation of a graph neural network framework, which consists of three TransformerConv layers for feature extraction and a multilayer perceptron module for classification. The benchmark results reflect that TransPTM has the competitive performance for non-histone acetylation site prediction over three state-of-the-art tools. It improves our comprehension on the PTM mechanism and provides a theoretical basis for developing drug targets for diseases. Moreover, the created PTM datasets fills the gap in non-histone acetylation site datasets and is beneficial to the related communities. The related source code and data utilized by TransPTM are accessible at https://www.github.com/TransPTM/TransPTM.

Advancements in Healthcare: Development of a Comprehensive Medical Information System with Automated Classification for Ocular and Skin Pathologies—Structure, Functionalities, and Innovative Development Methods

This article introduces a groundbreaking medical information system developed in Salesforce, featuring an automated classification module for ocular and skin pathologies using Google Teachable Machine. Integrating cutting-edge technology with Salesforce’s robust capabilities, the system provides a comprehensive solution for medical practitioners. The article explores the system’s structure, emphasizing innovative functionalities that enhance diagnostic precision and streamline medical workflows. Methods used in development are discussed, offering insights into the integration of Google Teachable Machine into the Salesforce framework. This collaborative approach is a significant stride in intelligent pathology classification, advancing the field of medical information systems and fostering efficient healthcare practices.

FIRCNet: Feature-based image reconstruction with classification learning for person re-identification

As an important task in computer vision, the person re-identification problem has two main approaches, identification and verification, in solving this problem. This paper proposed a network associating these two separate methods in training, including a reconstructive model. This design encompasses three core modules: a global feature extractor, a reconstructive decoder, and a classification network. Thus, the total loss function seamlessly integrates triplet loss, cross-entropy loss, and a reconstruction loss, optimizing the network for feature discriminability, accurate identity classification, and faithful image reconstruction concurrently. The global feature plays a pivotal role in both the training and testing phases, aiding in metric learning and ensuring a distinctive representation of identities. To train the decoder and the backbone network, which collectively forms the extractor, simultaneously, the reconstructive model as a part of a wide-range autoencoder can recover the original image from extracted features. Subsequently, the classification module classifies these features, assigning them to distinct person IDs, thereby aiding in precise identity recognition. Experimental results on three major re-id datasets give a demonstration of the notable improvement in the backbone network and the advantages of this approach compared to the state-of-the-art methods.

CREAD: A Classification-Restoration Framework with Error Adaptive Discretization for Watch Time Prediction in Video Recommender Systems

The watch time is a significant indicator of user satisfaction in video recommender systems. However, the prediction of watch time as a target variable is often hindered by its highly imbalanced distribution with a scarcity of observations for larger target values and over-populated samples for small values. State-of-the-art watch time prediction models discretize the continuous watch time into a set of buckets in order to consider the distribution of watch time. However, it is highly uninvestigated how these discrete buckets should be created from the continuous watch time distribution, and existing discretization approaches suffer from either a large learning error or a large restoration error. To address this challenge, we propose a Classification-Restoration framework with Error-Adaptive-Discretization (CREAD) to accurately predict the watch time. The proposed framework contains a discretization module, a classification module, and a restoration module. It predicts the watch time through multiple classification problems. The discretization process is a key contribution of the CREAD framework. We theoretically analyze the impacts of the discretization on the learning error and the restoration error, and then propose the error-adaptive discretization (EAD) technique to better balance the two errors, which achieves better performance over traditional discretization approaches. We conduct detailed offline evaluations on a public dataset and an industrial dataset, both showing performance gains through the proposed approach. Moreover, We have fully launched our framework to an online video platform, which resulted in a significant increase in users' video watch time by 0.29% through A/B testing. These results highlight the effectiveness of the CREAD framework in watch time prediction in video recommender systems.

FC-TFS-CGRU: A Temporal-Frequency-Spatial Electroencephalography Emotion Recognition Model Based on Functional Connectivity and a Convolutional Gated Recurrent Unit Hybrid Architecture.

The gated recurrent unit (GRU) network can effectively capture temporal information for 1D signals, such as electroencephalography and event-related brain potential, and it has been widely used in the field of EEG emotion recognition. However, multi-domain features, including the spatial, frequency, and temporal features of EEG signals, contribute to emotion recognition, while GRUs show some limitations in capturing frequency-spatial features. Thus, we proposed a hybrid architecture of convolutional neural networks and GRUs (CGRU) to effectively capture the complementary temporal features and spatial-frequency features hidden in signal channels. In addition, to investigate the interactions among different brain regions during emotional information processing, we considered the functional connectivity relationship of the brain by introducing a phase-locking value to calculate the phase difference between the EEG channels to gain spatial information based on functional connectivity. Then, in the classification module, we incorporated attention constraints to address the issue of the uneven recognition contribution of EEG signal features. Finally, we conducted experiments on the DEAP and DREAMER databases. The results demonstrated that our model outperforms the other models with remarkable recognition accuracy of 99.51%, 99.60%, and 99.59% (58.67%, 65.74%, and 67.05%) on DEAP and 98.63%, 98.7%, and 98.71% (75.65%, 75.89%, and 71.71%) on DREAMER in a subject-dependent experiment (subject-independent experiment) for arousal, valence, and dominance.

Hyperspectral Image Classification with the Orthogonal Self-Attention ResNet and Two-Step Support Vector Machine

Hyperspectral image classification plays a crucial role in remote sensing image analysis by classifying pixels. However, the existing methods require more spatial–global information interaction and feature extraction capabilities. To overcome these challenges, this paper proposes a novel model for hyperspectral image classification using an orthogonal self-attention ResNet and a two-step support vector machine (OSANet-TSSVM). The OSANet-TSSVM model comprises two essential components: a deep feature extraction network and an improved support vector machine (SVM) classification module. The deep feature extraction network incorporates an orthogonal self-attention module (OSM) and a channel attention module (CAM) to enhance the spatial–spectral feature extraction. The OSM focuses on computing 2D self-attention weights for the orthogonal dimensions of an image, resulting in a reduced number of parameters while capturing comprehensive global contextual information. In contrast, the CAM independently learns attention weights along the channel dimension. The CAM autonomously learns attention weights along the channel dimension, enabling the deep network to emphasise crucial channel information and enhance the spectral feature extraction capability. In addition to the feature extraction network, the OSANet-TSSVM model leverages an improved SVM classification module known as the two-step support vector machine (TSSVM) model. This module preserves the discriminative outcomes of the first-level SVM subclassifier and remaps them as new features for the TSSVM training. By integrating the results of the two classifiers, the deficiencies of the individual classifiers were effectively compensated, resulting in significantly enhanced classification accuracy. The performance of the proposed OSANet-TSSVM model was thoroughly evaluated using public datasets. The experimental results demonstrated that the model performed well in both subjective and objective evaluation metrics. The superiority of this model highlights its potential for advancing hyperspectral image classification in remote sensing applications.

Deep learning-based automated productivity monitoring for on-site module installation in off-site construction

Effectively monitoring and analyzing on-site module installation for modular integrated construction (MiC) is essential to properly coordinating the MiC process. In this study, the authors propose an automated productivity monitoring framework for on-site module installation operations consisting of three modules: object detection, activity classification, and productivity analysis. The object detection module detects mobile cranes and modules interacting with mobile cranes, and the activity classification module classifies module installation activities into five different activities by considering the spatiotemporal relationship between the detected objects. Finally, the productivity analysis module analyzes the productivity of the module installation process by utilizing the accumulated activity classification results over image frames. The proposed model achieves an average accuracy of 89% (hooking: 85.71%, lifting: 84.44%, positioning: 94.90%, returning: 83.09%, and idling: 96.87%) in classifying the five activities. The developed framework enables practitioners to measure the productivity of the on-site module installation process automatically. In addition, productivity data stored from diverse construction sites contribute to identifying progress-impeding factors and improving the productivity of the entire MiC process.

A Self-Adaptive Automatic Incident Detection System for Road Surveillance Based on Deep Learning.

We present an automatic road incident detector characterised by a low computational complexity for easy implementation in affordable devices, automatic adaptability to changes in scenery and road conditions, and automatic detection of the most common incidents (vehicles with abnormal speed, pedestrians or objects falling on the road, vehicles stopped on the shoulder, and detection of kamikaze vehicles). To achieve these goals, different tasks have been addressed: lane segmentation, identification of traffic directions, and elimination of unnecessary objects in the foreground. The proposed system has been tested on a collection of videos recorded in real scenarios with real traffic, including areas with different lighting. Self-adaptability (plug and play) to different scenarios has been tested using videos with significant scene changes. The achieved system can process a minimum of 80 video frames within the camera's field of view, covering a distance of 400 m, all within a span of 12 s. This capability ensures that vehicles travelling at speeds of 120 km/h are seamlessly detected with more than enough margin. Additionally, our analysis has revealed a substantial improvement in incident detection with respect to previous approaches. Specifically, an increase in accuracy of 2-5% in automatic mode and 2-7% in semi-automatic mode. The proposed classifier module only needs 2.3 MBytes of GPU to carry out the inference, thus allowing implementation in low-cost devices.

Spatial–Temporal Approach and Dataset for Enhancing Cloud Detection in Sentinel-2 Imagery: A Case Study in China

Clouds often cause challenges during the application of optical satellite images. Masking clouds and cloud shadows is a crucial step in the image preprocessing workflow. The absence of a thermal band in products of the Sentinel-2 series complicates cloud detection. Additionally, most existing cloud detection methods provide binary results (cloud or non-cloud), which lack information on thin clouds and cloud shadows. This study attempted to use end-to-end supervised spatial–temporal deep learning (STDL) models to enhance cloud detection in Sentinel-2 imagery for China. To support this workflow, a new dataset for time-series cloud detection featuring high-quality labels for thin clouds and haze was constructed through time-series interpretation. A classification system consisting of six categories was employed to obtain more detailed results and reduce intra-class variance. Considering the balance of accuracy and computational efficiency, we constructed four STDL models based on shared-weight convolution modules and different classification modules (dense, long short-term memory (LSTM), bidirectional LSTM (Bi-LSTM), and transformer). The results indicated that spatial and temporal features were crucial for high-quality cloud detection. The STDL models with simple architectures that were trained on our dataset achieved excellent accuracy performance and detailed detection of clouds and cloud shadows, although only four bands with a resolution of 10 m were used. The STDL models that used the Bi-LSTM and that used the transformer as the classifier showed high and close overall accuracies. While the transformer classifier exhibited slightly lower accuracy than that of Bi-LSTM, it offered greater computational efficiency. Comparative experiments also demonstrated that the usable data labels and cloud detection results obtained with our workflow outperformed the results of the existing s2cloudless, MAJA, and CS+ methods.