Image Recognition Research Articles

Consumer Usability Test of Mobile Food Safety Inquiry Platform Based on Image Recognition

Recently, as the types of imported food and the design of their packaging become more complex and diverse, digital recognition technologies such as barcodes, QR (quick response) codes, and OCR (optical character recognition) are attracting attention in order to quickly and easily check safety information (e.g., food ingredient information and recalls). However, consumers are still exposed to inaccurate and inconvenient situations because legacy technologies require dedicated terminals or include information other than safety information. In this paper, we propose a deep learning-based packaging recognition system which can easily and accurately determine food safety information with a single image captured through a smartphone camera. The detection algorithm learned a total of 100 kinds of product images and optimized YOLOv7 to secure an accuracy of over 95%. In addition, a new SUS (system usability scale)-based questionnaire was designed and conducted on 71 consumers to evaluate the usability of the system from the individual consumer’s perspective. The questionnaire consisted of three categories, namely convenience, accuracy, and usefulness, and each received a score of at least 77, which confirms that the proposed system has excellent overall usability. Moreover, in terms of task completion rate and task completion time, the proposed system is superior when it compared to existing QR code- or Internet-based recognition systems. These results demonstrate that the proposed system provides consumers with more convenient and accurate information while also confirming the sustainability of smart food consumption.

Automated evaluation of snow and ice conditions on airport pavement: A study based on image recognition and point cloud reconstruction

Automated evaluation of snow and ice conditions on airport pavement: A study based on image recognition and point cloud reconstruction

Advancements in Data Augmentation and Transfer Learning: A Comprehensive Survey to Address Data Scarcity Challenges

Abstract: Deep Learning (DL) models have demonstrated remarkable proficiency in image classification and recognition tasks, surpassing human capabilities. The observed enhancement in performance can be attributed to the utilization of extensive datasets. Nevertheless, DL models have huge data requirements. Widening the learning capability of such models from limited samples even today remains a challenge, given the intrinsic constraints of small datasets. The trifecta of challenges, encompassing limited labeled datasets, privacy, poor generalization performance, and the costliness of annotations, further compounds the difficulty in achieving robust model performance. Overcoming the challenge of expanding the learning capabilities of Deep Learning models with limited sample sizes remains a pressing concern even today. To address this critical issue, our study conducts a meticulous examination of established methodologies, such as Data Augmentation and Transfer Learning, which offer promising solutions to data scarcity dilemmas. Data Augmentation, a powerful technique, amplifies the size of small datasets through a diverse array of strategies. These encompass geometric transformations, kernel filter manipulations, neural style transfer amalgamation, random erasing, Generative Adversarial Networks, augmentations in feature space, and adversarial and meta- learning training paradigms. : Furthermore, Transfer Learning emerges as a crucial tool, leveraging pre-trained models to facilitate knowledge transfer between models or enabling the retraining of models on analogous datasets. Through our comprehensive investigation, we provide profound insights into how the synergistic application of these two techniques can significantly enhance the performance of classification tasks, effectively magnifying scarce datasets. This augmentation in data availability not only addresses the immediate challenges posed by limited datasets but also unlocks the full potential of working with Big Data in a new era of possibilities in DL applications.

Optimized binarization algorithm-based method for the image recognition and characterization of explosion damage in rock masses

The quantitative analysis of rock mass damage is crucial in fields such as engineering geology, disaster prevention, mining, geotechnical engineering, and structural engineering. With the advancement and application of noncontact measurement technologies and fractal theory, image-based damage identification methods are gaining increasing importance. This paper presents an optimized binarization algorithm for identifying and characterizing damage zones in granite explosion images. The method involves filtering, mathematical morphology operations, and connectivity recognition to effectively remove background noise while preserving clear boundaries of the damaged areas. It accurately captures the explosion damage in granite, both in terms of damage morphology and characteristic parameters. Additionally, the coefficient of agreement (COA) is introduced to quantitatively assess the accuracy of different methods in identifying damaged areas. The experimental results show that, compared with commonly used methods such as Otsu's method, Bernsen's algorithm, Niblack's algorithm, Sauvola's algorithm, and the K-means image clustering algorithm, the proposed method performs better in terms of identification accuracy and parameter agreement, achieving COA values near 1 across diverse experimental environments. Furthermore, the proposed method excels in handling uneven lighting, mitigating interference from rock surface textures and explosion carbonization zones, and demonstrates significant robustness in complex scenarios. The findings of this paper provide insights into the integration of engineering geology and computer vision technology. They offer valuable references for damage identification in excavation damage zones (EDZs), geological disaster evaluation, and structural damage warning systems.

Token Embeddings Augmentation benefits Parameter-Efficient Fine-Tuning under long-tailed distribution

Pre-trained vision-language models, particularly those utilizing CLIP, have advanced various visual tasks. Parameter-Efficient Fine-Tuning (PEFT) on such models is a mainstream trend for downstream tasks. Despite advancements, long-tailed distribution still hampers image recognition performance in current PEFT schemes. Therefore, this paper proposes Token Embeddings Augmentation (TEA) to tackle long-tailed learning under PEFT paradigm. Based on patch token semantic mining, TEA uncovers category-specific semantic details within patch tokens to enhance token embeddings, named Patch-based Embeddings Augmentation (PEA). Then, a Probability Gate (PG) strategy is designed to effectively enrich semantic information of tail categories using enhanced embeddings. A Token Embeddings Consistency (TEC) loss is further introduced to prioritize category semantic information within tokens. Extensive experiments on multiple long-tailed distribution datasets show that our method improves the performance of various PEFT methods with different classification loss functions, especially for tail categories. Our optimal approach achieves the state-of-the-art results on multiple datasets with negligible parameters or inference latency, thus enhancing the practicality of PEFT in long-tailed distributions.

Convolutional neural network augmented soft-sensor for autonomous microfluidic production of uniform bubbles

Microfluidics has emerged as a foundational process for creating highly uniform emulsions and bubbles. To enable integration of microfluidic platforms into industrial processes, achieving precise control over the size uniformity of microfluidic-generated bubbles and emulsions is crucial. Even if the external variables such as flow rates or pressures are kept constant, microfluidic processes can be easily disturbed by unknown factors that would substantially compromise the uniformity of resulting emulsions and bubbles. In this study, we introduce a two-step soft-sensor approach that combines a convolutional neural network (CNN) and an image recognition algorithm for feature extraction to detect both the flow regime and the size and uniformity of resulting bubbles. By using a CNN to detect flow regimes, our controller is able to restore the bubble-producing flow regime in response to disturbances. Beyond self-recovery, our controller actively adjusts to minimize errors, maintain setpoints, mitigate disturbances, and ensure system stability over extended periods. 99.2% of bubbles produced during an 8-hour period remain within 5% of the setpoint with our controller acting. By leveraging the soft sensor and artificial intelligence-assisted feedback control, our work presents a widely applicable approach for precise and automated control of microfluidics in diverse applications.

Paper-based triple-readout nanosensor for point-of-care detection of glucose in urine

Glucose detection is critical for diabetes diagnosis and management. This study aimed to develop an enzyme-free, cascade-based, triple-readout paper sensor utilizing bismuth (Bi)-based metal-organic frameworks/gold nanoparticles (Bi-BDC-NH2@Au) for detecting urine glucose. Herein, Bi-BDC-NH2@Au exhibited glucose oxidase-like activity and oxidized glucose to generate H2O2, which then quenched the blue fluorescence of FP@Bi-BDC-NH2@Au through an inner filter effect. Additionally, owing to its pronounced peroxidase-like activity, Bi-BDC-NH2@Au catalyzed H2O2 to produce ·OH, which oxidized colorless 3,3',5,5'-tetramethylbenzidine (TMB) to blue oxidized TMB (oxTMB), thereby increasing the system temperature owing to the excellent photothermal conversion properties of oxTMB. These mechanisms enabled the triple-readout of glucose levels. Quantification in the fluorescence and colorimetric modes was achieved through a Python-based image recognition algorithm that accurately read B/(R+G+B) values, and the photothermal mode relied on a portable infrared thermal imager to monitor temperature changes. The detection limits for the fluorescence, colorimetric, and photothermal modes were 4.2, 3.3, and 5.7μM, respectively. The self-calibration ability of the sensor across different modes markedly enhanced its detection accuracy and robustness. The developed sensor successfully has detected and quantified urine glucose, effectively distinguishing between healthy individuals and patients with diabetes, providing a convenient and efficient tool for diabetes management.

Optimization Strategy for Intelligent Traffic Monitoring Systems Based on Image Recognition

Optimization Strategy for Intelligent Traffic Monitoring Systems Based on Image Recognition

Photodual-Responsive Anthracene-Based 2D Covalent Organic Framework for Optoelectronic Synaptic Devices.

To facilitate the development of efficient neuromorphic perception and computation, it is crucial to explore optoelectronic synaptic devices that integrate perceptual and computational capabilities. Various materials such as oxide semiconductors, conjugated organic polymers, transition metal sulfides, perovskite materials, and metal nanoparticles, along with their composites, are utilized in constructing these devices. However, optoelectronic synaptic devices based on 2D covalent organic frameworks (COFs) is rarely reported. In this study, an anthracene-based 2D COF (COF-DaTp) film is prepared using a room-temperature interface-confined strategy and utilized it as the active layer in an optoelectronic synaptic device with an Al/COF-DaTp/ITO configuration. The device demonstrated dual optoelectronic modulation, exhibiting significant optoelectronic resistive switching in response to light pulses, achieving 32 photoconductive states. Moreover, it exhibited history-dependent memristive behavior in voltage scans and electrical pulses, with a comparable diversity of 32 conductive states. The photodual-responsive properties of the COF-DaTp-based synaptic device enable it to simultaneously perform optical sensing and basic image denoising and recognition tasks, significantly enhancing recognition accuracy and reducing the number of training epochs compared to datasets without noise mitigation. This work opens the door for the application of 2D COF-based optoelectronic synaptic devices in visual computational processing.

Imperial Study on Convolutional Layer in CNN for Data Mining

Convolutional Neural Networks (CNNs) have become a cornerstone in the field of data mining, particularly for tasks involving large-scale image and pattern recognition. This paper presents an imperial study on the efficacy of convolutional layers within CNN architectures in data mining applications. We analyze the impact of varying convolutional layer configurations, such as kernel size, stride, and depth, on performance metrics including accuracy, computational cost, and feature extraction quality. By conducting a series of experiments across different datasets, we explore how these configurations influence the model's ability to generalize and detect complex patterns in structured and unstructured data. Our findings indicate that optimizing the convolutional layers significantly improves the efficiency of CNNs for data mining tasks, offering insights into best practices for architectural design in such applications. This study provides valuable guidelines for leveraging CNNs in the realm of data mining, pushing the boundaries of automated data analysis and knowledge discovery

XAI-FruitNet: An explainable deep model for accurate fruit classification

In agricultural technology, precise fruit classification is essential yet challenging due to inherent interclass similarities and intra-class variabilities among fruit species. Despite their impressive performance, traditional deep learning models suffer from a lack of interpretability, which hampers their transparency and trustworthiness in practical applications. To address these issues, we present XAI-FruitNet, a novel hybrid deep learning architecture designed to enhance feature discrimination by integrating average and max pooling techniques. XAI-FruitNet, an optimized architecture for efficiency evaluated across the Fruits-360, Fruit Recognition, Fruit and Vegetables Image Recognition, and Dry Fruit datasets, consistently achieves over 97 % accuracy, surpassing existing state-of-the-art models and underscoring its remarkable generalization capability. A significant advancement of XAI-FruitNet is its built-in interpretability, which enhances the model's transparency and fosters trust among endusers. Through rigorous experimentation, we demonstrate that XAI-FruitNet advances state-of-the-art fruit classification accuracy and sets a new standard for explainable artificial intelligence (XAI) in agricultural applications. This hybrid approach ensures that stakeholders can rely on the classification outcomes' high performance and comprehensible nature, thereby offering a robust and trustworthy solution for modern agricultural needs.

Speech Emotion Recognition Using Transfer Learning: Integration of Advanced Speaker Embeddings and Image Recognition Models

Automatic Speech Emotion Recognition (SER) plays a vital role in making human–computer interactions more natural and effective. A significant challenge in SER development is the limited availability of diverse emotional speech datasets, which hinders the application of advanced deep learning models. Transfer learning is a machine learning technique that helps address this issue by utilizing knowledge from pre-trained models to improve performance on a new task in a target domain, even with limited data. This study investigates the use of transfer learning from various pre-trained networks, including speaker embedding models such as d-vector, x-vector, and r-vector, and image classification models like AlexNet, GoogLeNet, SqueezeNet, ResNet-18, and ResNet-50. We also propose enhanced versions of the x-vector and r-vector models incorporating Multi-Head Attention Pooling and Angular Margin Softmax, alongside other architectural improvements. Additionally, reverberation from the Room Impulse Response datasets was added to the speech utterances to diversify and augment the available data. Notably, the enhanced r-vector model achieved classification accuracies of 74.05% Unweighted Accuracy (UA) and 73.68% Weighted Accuracy (WA) on the IEMOCAP dataset, and 80.25% UA and 79.81% WA on the CREMA-D dataset, outperforming the existing state-of-the-art methods. This study shows that using cross-domain transfer learning is beneficial for low-resource emotion recognition. The enhanced models developed in other domains (for non-emotional tasks) can further improve the accuracy of SER.

A New Deep Learning Methodology for Alarm Supervision in Marine Power Stations.

Marine engineering officers operate and maintain the ship's machinery during normal navigation. Most accidents on board are related to human factors which, at the same time, are associated with the workload of the crew members and the working environment. The number of alarms is so high that, most of the time, instead of helping to prevent accidents, it causes more stress for crew members, which can result in accidents. Convolutional Neural Networks (CNNs) are being employed in the recognition of images, which depends on the quality of the images, the image recognition algorithm, and the very complex configuration of the neural network. This research study aims to develop a user-friendly image recognition tool that may act as a visual sensor of alarms adjusted to the particular needs of the ship operator. To achieve this, a marine engineering simulator was employed to develop an image recognition tool that advises marine engineering officers when they are conducting their maintenance activities, with the aim to reduce their stress as a work risk prevention tool. Results showed adequate accuracy for three-layer Convolutional Neural Networks and balanced data, and the use of external cameras stands out for user-friendly applications.

ARTIFICIAL INTELLIGENCE'S SIGNIFICANCE FOR THE NEXT GENERATION

It is similar to having a very intelligent computer that is capable of thought and learning. It all comes down to building machines that are capable of doing tasks that often call for human intelligence, such as language comprehension, image recognition, and decision-making. It's very remarkable. Imagine being able to learn and reason like a human brain is possible with a machine. That's the main purpose of artificial intelligence (AI). It's akin to endowing machines with the capacity for comprehension, decision-making, and even experience-based learning. KEY WORDS: Artificial Intelligence. Human, Process, Approach, Method etc.,

Comparison of deep learning-based models for detection of diseased trees using an image compression algorithm

The object of the research is the application of deep learning algorithms using an improved mathematical lossless image compression method for recognizing and identifying dead trees in aerospace images. The main problem that has been solved is the archiving of images due to their large volume on disk and the possibility of their further processing by deep learning methods such as convolutional and capsule neural networks, which have shown high efficiency and accuracy in image recognition and classification tasks using the proposed new image compression method. The article presents a comparative analysis of the performance of three YOLO (You Only Look Once) models with different types of architectures, such as YOLOv5, YOLOv7 and YOLOv8, to assess the effectiveness of their work for the task of recognizing aerospace tree images obtained from satellites, drones, and aircrafts. Comprehensive analysis of YOLO models presents that model YOLO v8 turned out to be most effective with a positive accuracy of 88.2 %, a recall of 77.4 %, and a mAP50 score of 87.2 %. Moreover, the average detection time was only 0.052 seconds for each image, even though the model size remains very small – 21.5 MB. These results suggest a much better usage of time and precise identification of dead trees, and classified targets with high efficiency. From the research, there is significant prospects of global forest management especially on forest reduction and protection of ecosystems through accurate assessment on the health of forestry. The proposed approach is universal and can be used in real life conditions, providing a good compromise of the speed, accuracy and resources required for forest monitoring and management

Artificial Intelligence (AI)-Based Radar Signal Processing and Radar Imaging

In recent years, artificial intelligence (AI), especially deep learning, has led to remarkable achievements in image recognition, speech recognition, autonomous driving and many other fields [...]

Effects of specimen thickness and fiber length on tensile and cracking behavior of UHPFRC: Uniaxial tensile test and micromechanical modeling

This study aims to investigate the effects of specimen thickness and fiber length on the tensile and cracking behaviors of ultra-high-performance fiber-reinforced concrete (UHPFRC). To this end, a uniaxial tensile test was conducted with three specimen thicknesses (30, 50, and 100 mm) and two fiber lengths (13 and 20 mm), and the fiber orientation, dispersion and specimen void were quantitatively evaluated based on image recognition. The test results indicated that fiber orientation was improved with the decreased specimen thickness and increased fiber length. Meanwhile, the initial cracking and peak stress, capacity to limit cracking were enhanced with the decreased specimen thickness. A modified prediction model considering the wall effect, flattening and squeezing effect was developed to predict the probability density function p(θ) of fiber orientation angle. Additionally, the uniformity factor μ2 was introduced to predict crack number, and the relationship between the μ2 and parameter ψ = (Vf × lf/df)/t was determined. Furthermore, a model was developed to convert the main crack width into uniaxial tensile strain. All models and relationships were validated using test data. A micromechanical model that considered the predicted p(θ) and conversion model was established to predict the uniaxial tensile response of UHPFRC, which was also successfully validated using test data.

Cross-coupled prompt learning for few-shot image recognition

Prompt learning based on large models shows great potential to reduce training time and resource costs, which has been progressively applied to visual tasks such as image recognition. Nevertheless, the existing prompt learning schemes suffer from either inadequate prompt information from a single modality or insufficient prompt interaction between multiple modalities, resulting in low efficiency and performance. To address these limitations, we propose a Cross-Coupled Prompt Learning (CCPL) architecture, which is designed with two novel components (i.e., Cross-Coupled Prompt Generator (CCPG) module and Cross-Modal Fusion (CMF) module) to achieve efficient interaction between visual and textual prompts. Specifically, the CCPG module incorporates a cross-attention mechanism to automatically generate visual and textual prompts, each of which will be adaptively updated using the self-attention mechanism in their respective image and text encoders. Furthermore, the CMF module implements a deep fusion to reinforce the cross-modal feature interaction from the output layer with the Image–Text Matching (ITM) loss function. We conduct extensive experiments on 8 image datasets. The experimental results verify that our proposed CCPL outperforms the SOTA methods on few-shot image recognition tasks. The source code of this project is released at: https://github.com/elegantTechie/CCPL.

LostNet: A smart way for lost and find.

The rapid population growth in urban areas has led to an increased frequency of lost and unclaimed items in public spaces such as public transportation, restaurants, and other venues. Services like Find My iPhone efficiently track lost electronic devices, but many valuable items remain unmonitored, resulting in delays in reclaiming lost and found items. This research presents a method to streamline the search process by comparing images of lost and recovered items provided by owners with photos taken when items are registered as lost and found. A photo matching network is proposed, integrating the transfer learning capabilities of MobileNetV2 with the Convolutional Block Attention Module (CBAM) and utilizing perceptual hashing algorithms for their simplicity and speed. An Internet framework based on the Spring Boot system supports the development of an online lost and found image identification system. The implementation achieves a testing accuracy of 96.8%, utilizing only 0.67 GFLOPs and 3.5M training parameters, thus enabling the recognition of images in real-world scenarios and operable on standard laptops.

Research on tomato disease image recognition method based on DeiT

Tomatoes, globally cultivated and economically significant, play an essential role in both commerce and diet. However, the frequent occurrence of diseases severely affects both yield and quality, posing substantial challenges to agricultural production worldwide. In China, where tomato cultivation is carried out on a large scale, disease prevention and identification are increasingly critical for enhancing yield, ensuring food safety, and advancing sustainable agricultural practices. As agricultural production scales and the demand for efficient methodologies grows, traditional disease recognition methods no longer meet current needs. The agricultural sector's move towards more modern and scalable production methods necessitates more effective and precise disease recognition technologies to support swift decision-making and timely preventive actions. To address these challenges, this paper proposes a novel tomato disease recognition method that integrates the data-efficient image transformers (DeiT) model with strategies like exponential moving average (EMA) and self-distillation, named EMA-DeiT. By leveraging deep learning technologies, this method significantly improves the accuracy of disease recognition. The enhanced EMA-DeiT model demonstrated exemplary performance, achieving a 99.6 % accuracy rate in identifying ten types of tomato leaf diseases within the PlantVillage public dataset and 98.2 % on the Dataset of Tomato Leaves, which encompasses six disease types. In generalization tests, it achieved 97.1 % accuracy on the PlantDoc dataset and 97.6 % on the Tomato-Village dataset. Utilizing the improved DeiT model, a comprehensive tomato disease recognition system was developed, featuring modules for image collection, disease detection, and information display. This system facilitates an integrated process from image collection to intelligent disease analysis, enabling agricultural workers to promptly understand and respond to disease occurrences. This system holds significant practical value for implementing precision agriculture and enhancing the efficiency of agricultural production.