Image Recognition Applications Research Articles

Artificial intelligence for image recognition in diagnosing oral and oropharyngeal cancer and leukoplakia

Visual diagnosis is one of the key features of squamous cell carcinoma of the oral cavity (OSCC) and oropharynx (OPSCC), both subsets of head and neck squamous cell carcinoma (HNSCC) with a heterogeneous clinical appearance. Advancements in artificial intelligence led to Image recognition being introduced recently into large language models (LLMs) such as ChatGPT 4.0. This exploratory study, for the first time, evaluated the application of image recognition by ChatGPT to diagnose squamous cell carcinoma and leukoplakia based on clinical images, with images without any lesion as a control group. A total of 45 clinical images were analyzed, comprising 15 cases each of SCC, leukoplakia, and non-lesion images. ChatGPT 4.0 was tasked with providing the most likely diagnosis based on these images in scenario one. In scenario two the image and the clinical history were provided, whereas in scenario three only the clinical history was given. The results and the accuracy of the LLM were rated by two independent reviewers and the overall performance was evaluated using the modified Artificial Intelligence Performance Index (AIPI. In this study, ChatGPT 4.0 demonstrated the ability to correctly identify leukoplakia cases using image recognition alone, while the ability to diagnose SCC was insufficient, but improved by including the clinical history in the prompt. Providing only the clinical history resulted in a misclassification of most leukoplakia and some SCC cases. Oral cavity lesions were more likely to be diagnosed correctly. In this exploratory study of 45 images of oral lesions, ChatGPT 4.0 demonstrated a convincing performance for detecting SCC only when the clinical history was added, whereas Leukoplakia was detected solely by image recognition. ChatGPT is therefore currently insufficient for reliable OPSCC and OSCC diagnosis, but further technological advancements may pave the way for the use in the clinical setting.

P-999. Novel diagnostic tools to enhance early detection of congenital Chagas disease

Abstract Background Chagas disease (CD), Trypanosoma cruzi infection, is a neglected vector-borne disease endemic in Latin America for which < 1% of cases have been treated. WHO estimates that ∼9,000 infected neonatal cases occur annually via vertical transmission—US CDC estimates up to ∼300 incident neonatal cases annually. Maternal and neonatal misdiagnosis are principally attributed to the paucity of accurate diagnostic tools for this complex chronic parasitic infection, with current guidelines requiring confirmation testing at >9 months of age—a significant barrier for vulnerable populations already experiencing healthcare access limitations. Methods A maternal-infant CD program was established in Sonsonate, El Salvador through a collaboration between flagship universities and the Ministry of Health. Employing machine learning methods, we developed a histology-based parasite image recognition application to aid in triaging of at-risk cases. Second, we developed an innovative diagnostic workflow integrating Bayesian statistics and novel digital PCR methods to validate a new diagnostic tool for enhanced disease detection among mothers-infants. Results Since March 2022, a 6% seropositivity was noted among parturition women. Maternal CD status was significantly associated with pregnancy complications and neonatal complications (Apgar < 9 at 1 min and at 5 mins, NICU admission, et al). A customized convolutional neural network design employed on bifurcated Giemsa-stained blood smears revealed >90% parasite detection accuracy. Lastly, McNemar’s statistic found significantly increased diagnostic capacity between our novel dPCR approach vs conventional qPCR. Specifically, dPCR was 2.74 times more likely to detect parasitic DNA in maternal blood than qPCR; 9.5 times more likely in placental tissue; 4.5 times more likely in cord tissue. Conclusion Congenital transmission is a growing international clinical concern. Our preliminary results provide compelling evidence for both the need of novel diagnostics and support for incorporation of study-developed cutting-edge diagnostics to aid in timely clinical management of congenital CD. Disclosures All Authors: No reported disclosures

Technological Development and Application of Image Recognition - License Plate Recognition

Technological Development and Application of Image Recognition - License Plate Recognition

Soft Artificial Synapse Electronics.

Soft electronics, known for their bendable, stretchable, and flexible properties, are revolutionizing fields such as biomedical sensing, consumer electronics, and robotics. A primary challenge in this domain is achieving low power consumption, often hampered by the limitations of the conventional von Neumann architecture. In response, the development of soft artificial synapses (SASs) has gained substantial attention. These synapses seek to replicate the signal transmission properties of biological synapses, offering an innovative solution to this challenge. This review explores the materials and device architectures integral to SAS fabrication, emphasizing flexibility and stability under mechanical deformation. Various architectures, including floating-gate dielectric, ferroelectric-gate dielectric, and electrolyte-gate dielectric, are analyzed for effective weight control in SASs. The utilization of organic and low-dimensional materials is highlighted, showcasing their plasticity and energy-efficient operation. Furthermore, the paper investigates the integration of functionality into SASs, particularly focusing on devices that autonomously sense external stimuli. Functionalized SASs, capable of recognizing optical, mechanical, chemical, olfactory, and auditory cues, demonstrate promising applications in computing and sensing. A detailed examination of photo-functionalized, tactile-functionalized, and chemoreception-functionalized SASs reveals their potential in image recognition, tactile sensing, and chemosensory applications, respectively. This study highlights that SASs and functionalized SAS devices hold transformative potential for bioelectronics and sensing for soft-robotics applications; however, further research is necessary to address scalability, long-time stability, and utilizing functionalized SASs for prosthetics and invivo applications through clinical adoption. By providing a comprehensive overview, this paper contributes to the understanding of SASs, bridging research gaps and paving the way toward transformative developments in soft electronics, biomimicking and biointegrated synapse devices, and integrated systems.

Automatic defect detection of pipelines based on improved OFG-YOLO algorithm

Underground metal pipelines are a key mode of long-distance transportation for oil, gas, and water owing to their low transport costs and high safety standards. Nevertheless, these pipelines are vulnerable to defects that gradually emerge during operation, potentially impairing their functionality and maintenance. The timely detection and maintenance of defects in metal pipelines is therefore of great significance. Although pipeline defects are presently detected via closed-circuit television (CCTV) and ultrasonic time-of-flight diffusion technology (TOFD), these methods depend upon engineering experience and are not sufficiently intuitive. In contrast, the application of image recognition for pipeline defect detection has demonstrated considerable potential. This paper presents an automatic metal pipeline defect detection method based on the YOLO architecture, which was modified by enhancing the convolutional, feature fusion, and regression layers, as well as the spatial pyramid pooling module. Our experimental results demonstrate that the enhanced model outperforms mainstream algorithms in terms of various evaluation indicators.

Resnet for Blood Sample Detection: A Study on Improving Diagnostic Accuracy

Automated blood cell analysis plays a crucial role in medical diagnostics, enabling rapid and accurate assessment of a patient's health status. In this paper, we provide a unique technique for detecting and classifying WBCs,RBCs, and platelets inside blood smear pictures using ResNet (Residual Neural Network), a deep learning architecture. Because of its capacity to efficiently train very deep neural networks while minimizing the vanishing gradient problem, the ResNet architecture has exhibited excellent performance in a variety of image recognition applications. Leveraging the power of ResNet, we developed a multi-class classification model capable of distinguishing between WBCs, RBCs, and platelets within microscopic images of blood smears. Our methodology involved preprocessing the blood smear images to enhance contrast and remove noise, followed by image segmentation to isolate individual blood cells and platelets. The segmented images were then used to train and fine-tune a ResNet model, utilizing a large annotated dataset of labeled blood cell images. The trained model exhibited remarkable accuracy in identifying and classifying different blood cell types, even in the presence of overlapping cells or artifacts. We extensively tested our suggested technique, on a range of blood smear images to evaluate its performance. The findings demonstrated that ResNet effectively identifies and categorizes WBCs, (RBCs) and platelets. When compared to methods our approach showcased superior accuracy, robustness and generalization capabilities. After training the model with the Resnet algorithm we got 92% of Accuracy

Dynamic hybrid visual-thermal multimodal perception neuromorphic devices based on defect modulation of electrospun nanofibers

Abstract Neuromorphic devices, inspired by the intricate architecture of the human brain, have garnered recognition for their prodigious computational speed and sophisticated parallel computing capabilities. Vision, the primary mode of external information acquisition in living organisms, has garnered substantial scholarly interest. Notwithstanding numerous studies simulating the retina through optical synapses, their applications remain circumscribed to single-mode perception. Moreover, the pivotal role of temperature, a fundamental regulator of biological activities, has regrettably been relegated to the periphery. To address these limitations, we proffer a neuromorphic device endowed with multimodal perception, grounded in the principles of light-modulated semiconductors. This device seamlessly accomplishes dynamic hybrid visual and thermal multimodal perception, featuring temperature-dependent PPF properties and adaptive storage. Crucially, our meticulous examination of transfer curves, capacitance-voltage (C-V) tests, and noise measurements provides insights into interface and bulk defects, elucidating the physical mechanisms underlying adaptive storage and other functionalities. Additionally, the device demonstrates a variety of synaptic functionalities, including filtering properties, Ebbinghaus curves, and memory applications in image recognition. Surprisingly, the digital recognition rate achieves a remarkable value of 98.8%. These discernments furnish crucial insights for the prospective evolution of intricate neuromorphic systems.

Interaction of Asymmetric Adaptive Network Structures and Parameter Balance in Image Feature Extraction and Recognition

To better process irregular sample images for their image feature extraction and recognition, this essay proposes asymmetric adaptive neural network (AACNN) structures, including dual structures of an adaptive image feature extraction network (AT-CNN) and adaptive image recognition network (AT-ACNN). They both comprise an Adaptive Transform (AT) module and a deep learning network, but the ACNN comprises pixel-adaptive convolutional (PAC) kernels that CNN does not have, reflecting the asymmetry of these network structures. Structural analysis and comparative testing experiments indicated that the proposed method is more appropriate and effective for dealing with irregular sample images with different sizes and views, mainly focusing on their feature extraction accuracy and image recognition efficiency. The proposed method constructs the interaction between asymmetric dual network structures, essential in improving model performance and efficiency. It specifically manifests that the PAC kernels in an ACNN resolves the problem of content-agnostic convolution in image recognition by learning image features from a pre-trained CNN. On the other hand, it improves image recognition efficiency by using feature maps extracted from the pre-trained CNN to train the classifiers in the ACNN. We also found that parameter balance was essential in adaptive neural network structure for better performance in further testing experiments. When setting the Dropout layer parameter at 0.5 and the iteration number was 32, the proposed model achieved adequate recognition accuracy and efficiency. Smaller parameters affect model performance, but more extensive parameters significantly increase computational burden and loss. Comparative testing experiments fully validated its superiority compared with traditional methods based on CNNs. Using traditional carving patterns from Anhui Province as an example, we conducted image recognition and feature graphic application under ideal parameter balance conditions and thereby demonstrated the practicality and value of the proposed method.

Machine vision-based recognition of safety signs in work environments.

The field of image recognition is extensively researched, with applications addressing numerous challenges posed by the scientific community. Notably among these challenges are those related to individual safety. This article presents a system designed for the application of image recognition in the realm of Occupational Risk Prevention-a concern of paramount importance due to the imperative of preventing workplace accidents as falls, collisions, or other types of accidents for the benefit of both workers and enterprises. In this study, convolutional neural networks are employed due to their exceptional efficacy in image recognition. Leveraging this technology, the focus is on the recognition of safety signs used in Occupational Risk Prevention. The primary objective is to enable the recognition of these signs regardless of their orientation or potential degradation, phenomena commonly observed due to regular exposure to environmental elements or deliberate defacement. The results of this research substantiate the feasibility of integrating this technology into devices capable of promptly alerting individuals to potential risks. However, to improve classification capabilities, especially for highly degraded or complex images, a larger and more diverse data set might be needed, including real-world images that introduce greater entropy and variability. Implementing such a system would provide workers and companies with a proactive measure against workplace accidents, thereby enhancing overall safety in occupational environments.

Research on Intelligent Screening of International Communication Information and National and Regional Communication Strategies Based on Deep Learning Algorithm

In the context of globalization, the importance of international communication has become increasingly prominent. However, in the face of massive international communication information, how to effectively select and formulate communication strategies has become a big challenge. The traditional manual screening method is not only time-consuming and labor-intensive but also easy to be affected by subjective factors. This study aims to address the efficiency and accuracy issues in current international communication information screening and dissemination strategy formulation through the application of deep learning algorithms. It is of great theoretical and practical value to use deep learning algorithms to conduct intelligent screening of international communication information and research on national and regional communication strategies. To explore the application of deep learning algorithms in the intelligent screening of international communication information and the research of national and regional communication strategies, first, we will introduce the basic principles and methods of deep learning, as well as its applications in natural language processing (NLP), image recognition, and other fields. Then, we will explore how to use deep learning algorithms for intelligent screening of international dissemination information, including information classification, clustering, topic modeling, etc. Combined with practical cases, the application of deep learning algorithms in national and regional communication strategies will be studied, including communication content generation, communication channel selection, and communication effect evaluation.

Bio-Inspired Neuromorphic Sensory Systems from Intelligent Perception to Nervetronics.

Inspired by the extensive signal processing capabilities of the human nervous system, neuromorphic artificial sensory systems have emerged as a pivotal technology in advancing brain-like computing for applications in humanoid robotics, prosthetics, and wearable technologies. These systems mimic the functionalities of the central and peripheral nervous systems through the integration of sensory synaptic devices and neural network algorithms, enabling external stimuli to be converted into actionable electrical signals. This review delves into the intricate relationship between synaptic device technologies and neural network processing algorithms, highlighting their mutual influence on artificial intelligence capabilities. This study explores the latest advancements in artificial synaptic properties triggered by various stimuli, including optical, auditory, mechanical, and chemical inputs, and their subsequent processing through artificial neural networks for applications in image recognition and multimodal pattern recognition. The discussion extends to the emulation of biological perception via artificial synapses and concludes with future perspectives and challenges in neuromorphic system development, emphasizing the need for a deeper understanding of neural network processing to innovate and refine these complex systems.

A study of machine learning applications in healthcare

Abstract. Machine learning, a sub-field of artificial intelligence, has numerous algorithms with applications in image recognition and classification, natural language processing, data prediction, medical diagnosis, and more, achieving significant results. Nowadays, the application of many algorithms of machine learning in the medical field has become a research hotspot and attracted wide attention from society, which plays a key role in disease screening, disease prediction, and assisted diagnosis of diseases. In this paper, we have developed a new approach to the medical field that can be applied in the healthcare industry by using technologies such as k-nearest neighbor (KNN), recurrent neural network (RNN), plain Bayesian (NB), decision tree (DTT), random forest (RF), deep convolutional adversarial network (DCAN), generative adversarial network (GAN), deep convolutional generative adversarial network (DCGAN),deep convolutional neural network (CNN), and support vector machine (SVM). By referring to the existing related literature and application practice results, this paper provides a systematic introduction to the application and research of machine learning in the medical field. This paper also analyses and describes the challenges and difficulties in the implementation of machine learning in the medical field.

Application of AI-Driven Medical Image Recognition in Precision Medicine and Healthcare

Abstract. Artificial Intelligence (AI) has been playing an important role in medical image recognition, especially in precision medicine and health management, where AI can improve the accuracy and efficiency of diagnosis by utilizing advanced technologies such as deep learning. This paper explores in detail the application of AI techniques in the early diagnosis of cancer, image analysis of cardiovascular diseases, and assessment of neuropsychiatric disorders. We also identify challenges in technical standards, data security and privacy protection, and interdisciplinary collaboration, and emphasize the importance of establishing effective technical standards and evaluation systems, as well as fostering collaboration among experts in various fields.

Research on image recognition and processing application technology of unmanned vehicle based on deep learning

Abstract. Image recognition technology is critically important in various fields, including the rapidly advancing sector of autonomous vehicles. As one of the core components enabling driverless cars to perceive their environment and make informed decisions, image recognition has seen significant advancements due to deep learning. This paper focuses on the application of deep learning in image recognition for self-driving cars and explores its implications for the future of autonomous driving technology. To begin with, this paper examines the empirical evaluation of deep learning models in highway driving scenarios. By employing Convolutional Neural Networks (CNN), these models achieve high detection rates and superior accuracy in recognizing vehicles and lanes. The robustness of these models is tested under varying weather conditions and times, demonstrating their effectiveness compared to classical computer vision techniques. Next, the paper discusses radar-camera fusion technology, highlighting different fusion strategies such as data-level, feature-level, object-level, and hybrid-level. The findings suggest that while feature-level fusion excels in detecting small objects in complex scenes, hybrid-level fusion is optimal for diverse driving situations. This section provides valuable insights into the integration of multimodal data for improved object recognition and semantic segmentation. After discussing fusion technologies, the paper finally reviews the security challenges posed by adversarial attacks on deep learning-based unmanned systems.

Exploring Large Language Models and the Metaverse for Urologic Applications: Potential, Challenges, and the Path Forward.

The metaverse, a 3-dimensional digital platform that enables users to interact and engage in realistic virtual activities beyond time and space limitations, has garnered significant investment across industries, particularly in healthcare. In the medical field, the metaverse shows promise as a digital therapeutic platform to enhance interaction between medical professionals and patients. Concurrently, generative artificial intelligence, especially large language models, is being integrated into healthcare for applications in data analysis, image recognition, and natural language processing. In urology, large language models (LLMs) support are increasingly used in urology for tasks such as image diagnosis, data processing, patient education, and treatment assistance in order to provide significant support in clinical settings. By combining LLMs with the immersive capabilities of the metaverse, new possibilities emerge to improve urologic treatment in areas that require consistent treatments, habit formation, and long-term management. This paper reviews current research and applications of LLMs in urology, discusses the challenges associated with their use including data quality, bias, security, and ethical issues, and explores the need for regulatory standards. Furthermore, it highlights the potential of a metaverse-based digital platform to improve urologic care and streamline information exchange to maximize the benefits of this integrated approach in future healthcare applications.

Performance Optimization of Convolutional Neural Networks in Image Classification

With the development of deep learning, CNNs have become mainstream in image classification. However, CNN performance is affected by factors like loss function choice and optimization algorithm. This paper aims to improve CNN performance in image classification by optimizing SGD. Firstly, we analyze different loss functions and propose a suitable optimization strategy. Second, we explore SGD and its variants' roles in CNN training, focusing on accelerating convergence and improving accuracy. Experiments show the proposed scheme effectively enhances CNN performance, providing a reliable solution for image classification. This study supports CNN applications in image recognition and lays a foundation for exploring more efficient optimization methods.

An improved transfer learning algorithm using integration method and its application in image segmentation and recognition

An improved transfer learning algorithm using integration method and its application in image segmentation and recognition

Enhancing computer image recognition with improved image algorithms

Advances in computer image recognition have significantly impacted many industries, including healthcare, security and autonomous systems. This paper aims to explore the potential of improving image algorithms to enhance computer image recognition. Specifically, we will focus on regression methods as a means to improve the accuracy and efficiency of identifying images. In this study, we will analyze various regression techniques and their applications in computer image recognition, as well as the resulting performance improvements through detailed examples and data analysis. This paper deals with the problems related to visual image processing in outdoor unstructured environment. Finally, the heterogeneous patterns are converted into the same pattern, and the heterogeneous patterns are extracted from the fusion features of data modes. The simulation results show that the perception ability and recognition ability of outdoor image recognition in complex environment are improved.

The effect of mobile image recognition-based augmented reality application on students’ acquisition of science processing skills and motivations to learn science

ABSTRACT With the development of mobile phone applications with innovative features, the application of mobile recognition technology, such as a mobile image recognition-based (AR) application, is becoming increasingly widespread. Mobile Image Recognition-Based AR connect nearby objects or other physical targets with digital information. The purpose of this study is to investigate the effects of using an image recognition-based AR application on students’ acquisition of science process skills and their motivation toward science learning. The study involved 56 students enrolled in two sections of tenth-grade female students in a private school in Jordan. Each section was randomly assigned to either the control group or the experimental group. The experimental group studied a unit from the biology textbook with image-based recognition content. In contrast, the control group studied the same unit using the conventional teaching method based on the traditional textbook content. The researchers administered a science both groups a science process skills test and a motivation toward learning science scale. Results showed that participants who used a mobile image recognition application to learn science had higher acquisition of all science skills and a higher motivation toward learning science than those who used the conventional learning method.

Image Data Acquisition and Classification of Vannamei Shrimp Cultivation Results Based on Deep Learning

This research aimed to employ deep learning techniques to address the classification of Litopenaeus vannamei cultivation results in land ponds and tarpaulin ponds. Despite their similar appearance, distinguishable differences exist in various aspects such as color, shape, size, and market price between the two cultivation methods, often leading to consumer confusion and potential exploitation by irresponsible sellers. To mitigate this challenge, the research proposed a classification method utilizing two Convolutional Neural Network (CNN) architectures: Visual Geometry Group-16 (VGG-16) and Residual Network-50 (ResNet-50), renowned for their success in various image recognition applications. The dataset comprised 2,080 images per class of vannamei shrimp from both types of ponds. Augmentation techniques enhanced the dataset’s diversity and sample size, reinforcing the model’s ability to discern shrimp morphology variations. Experiments were conducted with learning rates of 0.001 and 0.0001 on the Stochastic Gradient Descent (SGD) and Adaptive Moment Estimation (ADAM) optimizers to evaluate their effectiveness in model training. The VGG-16 and ResNet-50 models were trained with a learning rate parameter of 0.0001, leveraging the flexibility and reasonable control provided by the SGD optimizer. Lower learning rate values were chosen to prevent overfitting and enhance training stability. The model evaluation demonstrated promising results, with both architectures achieving 100% accuracy in classifying vannamei shrimp from soil ponds and tarpaulin ponds. Furthermore, experimental findings highlight the superiority of using SGD with a learning rate of 0.0001 over 0.001 on both architectures, underscoring the significant impact of optimizer and learning rate selection on model training effectiveness in image classification tasks.