Search

Ripeness classification is crucial in ensuring the quality and marketability of avocados. This paper aims to develop the Multi-Channel Hybrid Deep Neural Networks (MCHDNN) model between Visual Geometry Group 16 (VGG16) and EfficientNetB0 architectures, tailored explicitly for avocado ripeness classification in five classes: firm, breaking, ripe, overripe, and rotten. Each feature extracted is concatenated in an early fusion-based to classify the ripeness. The image dataset used for each avocado fruit was captured from six sides: front, back, left, right, bottom, and pedicel to provide a multi-channel input image in of a Convolution Neural Network (CNN). The results showed that the developed fine-tuned MCHDNN had an accuracy of 94.10% in training, 90.13% in validation, and 90.18% in testing. In addition, when considering individual class classification in the confusion matrix of the training set, it was found that the 'ripe' class had the highest accuracy of 94.58%, followed by the 'firm' and 'rotten' classes with 94.50% and 93.75% accuracy, respectively. Moreover, compared with the single-channel model, the fine-tuned MCHDNN model performs 7.70% more accurately than the fine-tuned VGG16 model and 7.77% more accurately than the fine-tuned EfficientNetB0 model. [ABSTRACT FROM AUTHOR]

The study of natural disasters is a crucial field that involves analyzing the occurrence, impact, and aftermath of various natural hazards that can cause significant harm to communities and the environment. Efficient waste management and environmental protection require proper classification of waste. Analyzing natural disasters and categorizing waste can be a time-consuming task, and conventional methods often struggle with it. However, a new approach called Visual Geometry Group with Federated Learning (VGG-FL) has been introduced to address these challenges. This methodology uses the golden search optimization (GSO) algorithm for feature selection and leverages VGG with federated learning for feature extraction and classification. To test the effectiveness of this method, a disaster image dataset was used to train the VGG-FL model. The results showed that the VGG-FL model attained exceptional accuracy in discerning and categorizing various disaster scenarios. The waste classification dataset simultaneously trains the VGG-FL model to categorize waste based on its characteristics and potential hazards. To measure the model's performance, several evaluation metrics such as accuracy, specificity, precision, F1-score, and recall are utilized to assess the effectiveness of the proposed VGG-FL method. These results are then compared with existing methodologies. The VGG-FL method performs exceptionally well, achieving 98.52% accuracy, 97.48% precision, 97.83% recall, 97.58% F1-score, and 97.12% specificity. These experimental findings demonstrate the efficacy of the VGG-FL method in analyzing natural disasters and classifying waste materials. [ABSTRACT FROM AUTHOR]

Abstract The cultivation of grapes encounters various challenges, such as the presence of pests and diseases, which have the potential to considerably diminish agricultural productivity. Plant diseases pose a significant impediment, resulting in diminished agricultural productivity and economic setbacks, thereby affecting the quality of crop yields. Hence, the precise and timely identification of plant diseases holds significant importance. This study employs a Convolutional neural network (CNN) with and without data augmentation, in addition to a DCNN Classifier model based on VGG16, to classify grape leaf diseases. A publicly available dataset is utilized for the purpose of investigating diseases affecting grape leaves. The DCNN Classifier Model successfully utilizes the strengths of the VGG16 model and modifies it by incorporating supplementary layers to enhance its performance and ability to generalize. Systematic evaluation of metrics, such as accuracy and F1-score, is performed. With training and test accuracy rates of 99.18 and 99.06%, respectively, the DCNN Classifier model does a better job than the CNN models used in this investigation. The findings demonstrate that the DCNN Classifier model, utilizing the VGG16 architecture and incorporating three supplementary CNN layers, exhibits superior performance. Also, the fact that the DCNN Classifier model works well as a decision support system for farmers is shown by the fact that it can quickly and accurately identify grape diseases, making it easier to take steps to stop them. The results of this study provide support for the reliability of the DCNN classifier model and its potential utility in the field of agriculture.

The cultivation of grapes encounters various challenges, such as the presence of pests and diseases, which have the potential to considerably diminish agricultural productivity. Plant diseases pose a significant impediment, resulting in diminished agricultural productivity and economic setbacks, thereby affecting the quality of crop yields. Hence, the precise and timely identification of plant diseases holds significant importance. This study employs a Convolutional neural network (CNN) with and without data augmentation, in addition to a DCNN Classifier model based on VGG16, to classify grape leaf diseases. A publicly available dataset is utilized for the purpose of investigating diseases affecting grape leaves. The DCNN Classifier Model successfully utilizes the strengths of the VGG16 model and modifies it by incorporating supplementary layers to enhance its performance and ability to generalize. Systematic evaluation of metrics, such as accuracy and F1-score, is performed. With training and test accuracy rates of 99.18 and 99.06%, respectively, the DCNN Classifier model does a better job than the CNN models used in this investigation. The findings demonstrate that the DCNN Classifier model, utilizing the VGG16 architecture and incorporating three supplementary CNN layers, exhibits superior performance. Also, the fact that the DCNN Classifier model works well as a decision support system for farmers is shown by the fact that it can quickly and accurately identify grape diseases, making it easier to take steps to stop them. The results of this study provide support for the reliability of the DCNN classifier model and its potential utility in the field of agriculture. [ABSTRACT FROM AUTHOR]

Human emotion recognition is crucial in various technological domains, reflecting our growing reliance on technology. Facial expressions play a vital role in conveying and preserving human emotions. While deep learning has been successful in recognizing emotions in video sequences, it struggles to effectively model spatio-temporal interactions and identify salient features, limiting its accuracy. This research paper proposed an innovative algorithm for facial expression recognition which combined a deep learning algorithm and dynamic texture methods. In the initial phase of this study, facial features were extracted using the Visual-Geometry-Group (VGG19) model and input into Long-Short-Term-Memory (LSTM) cells to capture spatio-temporal information. Additionally, the HOG-HOF descriptor was utilized to extract dynamic features from video sequences, capturing changes in facial appearance over time. Combining these models using the Multimodal-Compact-Bilinear (MCB) model resulted in an effective descriptor vector. This vector was then classified using a Support Vector Machine (SVM) classifier, chosen for its simpler interpretability compared to deep learning models. This choice facilitates better understanding of the decision-making process behind emotion classification. In the experimental phase, the fusion method outperformed existing state-of-the-art methods on the eNTERFACE05 database, with an improvement margin of approximately 1%. In summary, the proposed approach exhibited superior accuracy and robust detection capabilities. [ABSTRACT FROM AUTHOR]

Vascular Dementia is a severe disease that results from dead nerve cells’ accumulation in blood vessels. This affects the blood flow and impairs memory and decision-making abilities. Machine learning and deep learning have been used in detecting this disease. Nevertheless, their accuracy has been inconsistent, explaining why their utilization in diagnosing patients has led to poor performance. We developed several transfer learning architectures that improve classification accuracy and diagnosis performance in assessing vascular dementia. The process first entails the preprocessing of the dataset where a random selection ensures data representation is balanced. We used a dataset containing resting-state fMRI scans to split training, testing, and validation into 80%, 10%, and 10%. We employ different Convolutional Neural Network architectures such as VGG16, VGG19, DenseNet121, and InceptionResNetV2 to enhance classification. To, enhance these, we incorporated Rectified Linear Unit and leaky activation functions for the training phase to counteract problems associated with vanishing gradient common in deep learning tasks. Our methodology ensures effective information flow throughout different layers, which is essential for a divergent information hierarchy in medical information. As such, the specifics show that our approach achieved 84.67% in accuracy in the multi-classification, which is better than the current state-of-the-art research in the same field. Therefore, the result shows that transfer learning-based approaches are suitable when combined with strategic pre-processing and activation functions in improving the diagnosis of vascular dementia using MRI images.

Wireless capsule endoscopy is one of the diagnostic methods used to record the video of the gastrointestinal tract. The endoscopy capsule stays in the digestive system for at least eight hours. It is difficult for gastroenterologists to examine such a lengthy video and identify the ailment. Convolutional neural networks (CNN) are a powerful solution to several computer vision problems. CNN can speed up the reviewing time of the recorded video by classifying video frames into various categories. The primary emphasis of this research paper is to examine and evaluate the performance of three different CNN architectures-VGG, inception, and MobileNet-in classifying the disease. Experimental results demonstrate that MobileNetV2's accuracy is 91%, whereas InceptionV3 and VGG16 have an accuracy of 94% which is better than the accuracy of MobileNetV3. However, MobileNeV2 performed relatively better than the other CNN models in terms of computational time and cost. The model's F-score, precision, and recall values are computed and compared also. [ABSTRACT FROM AUTHOR]

In 2019, Corona Virus Disease (COVID)-19 has created an important impact on people's health and economy because of its rapid spreading. Therefore, the earlier detection and diagnosis for handling the further spread is regarded as a significant factor. Computed Tomography (CT) images of the lungs are utilised for the detection process of COVID-19 infection. Chest X-Ray (CXR) and CT serves as the major source for the identification of COVID-19. Some of the classifiers have shown promising outcomes as well as better performance over various medical imaging applications and computer vision processes. Due COVID pandemic, researchers have utilised standard techniques for accurately identifying the coronavirus infection in lung images. Here, a new COVID-19 detection approach has been implemented with the adoption of ensemble classifiers to recognise the occurrence of COVID-19 infection at the starting phase. The required CT and the CXR images are garnered from publicly presented online sources. Meanwhile, the acquired images undergo the process of pre-processing using filtering methods and histogram equalisation techniques in order to ignore the unnecessary artifacts are presented in the images. Then, the detection of COVID-19 is made possible through Ensemble Classifiers (EC) such as Multi-scale Convolutional Neural Networks (CNN), Residual Attention Networks (RAN), Visual Geometry Group (VGG), and Long Short Term Memory (LSTM), where the variables within the machine learning strategies are tuned by Modified Mexican Axolotl Shuffled Frog-Leaping Optimization (MMASFLO). The efficacy of the newly developed ensemble classifier-based COVID-19 detection approaches will be tested through various benchmark datasets, and is tested with several existing approaches and conventional COVID-19 detection approaches. [ABSTRACT FROM AUTHOR]

Analyzing and understanding emotion can help in various aspects, such as realizing one's attitude, behavior, etc. By understanding one's emotions, one's mental health state can be calculated, which can help in the medical field by classifying whether one is mentally stable or not. Facial Recognition is one of the many fields of computer vision that utilizes convolutional networks or Conv Nets to perform, train, and learn. Conv Nets and other machine learning algorithms have evolved to adapt better to larger datasets. One of the advancements in Conv Nets and machines is the introduction of various Conv architectures like VGGNet. Thus, this study will present a mental health state classification approach based on facial emotion recognition. The methodology comprises several interconnected components, including preprocessing, feature extraction using Principal Component Analysis (PCA) and VGGNet, and classification using Support Vector Machines (SVM) and Multilayer Perceptron (MLP). The FER2013 dataset tests multiple models' performances, and the best model is employed in the mental health state classification. The best model, which combines Visual Geometry Group Network (VGGNet) feature extraction with SVM classification, achieved an accuracy of 66%, demonstrating the effectiveness of the proposed methodology. By leveraging facial emotion recognition and machine learning techniques, the study aims to develop an effective method. [ABSTRACT FROM AUTHOR]

Human emotion recognition is crucial in various technological domains, reflecting our growing reliance on technology. Facial expressions play a vital role in conveying and preserving human emotions. While deep learning has been successful in recognizing emotions in video sequences, it struggles to effectively model spatio-temporal interactions and identify salient features, limiting its accuracy. This research paper proposed an innovative algorithm for facial expression recognition which combined a deep learning algorithm and dynamic texture methods. In the initial phase of this study, facial features were extracted using the Visual-Geometry-Group (VGG19) model and input into Long-Short-Term-Memory (LSTM) cells to capture spatio-temporal information. Additionally, the HOG-HOF descriptor was utilized to extract dynamic features from video sequences, capturing changes in facial appearance over time. Combining these models using the Multimodal-Compact-Bilinear (MCB) model resulted in an effective descriptor vector. This vector was then classified using a Support Vector Machine (SVM) classifier, chosen for its simpler interpretability compared to deep learning models. This choice facilitates better understanding of the decision-making process behind emotion classification. In the experimental phase, the fusion method outperformed existing state-of-the-art methods on the eNTERFACE05 database, with an improvement margin of approximately 1%. In summary, the proposed approach exhibited superior accuracy and robust detection capabilities.

Brain cancer is one of the most deadly forms of cancer today, and its timely and accurate diagnosis can significantly impact the patient's quality of life. A computerized tomography scan (CT) and magnetic resonance imaging (MRI) of the brain is required to diagnose this condition. In the past, several methods have been proposed as a means of diagnosing brain tumors through the use of medical images. However, due to the similarity between tumor tissue and other brain tissues, these methods have not proven to be accurate. A novel method for diagnosing brain tumors using MRI and CT scan images is presented in this paper. An architecture based on deep learning is used to extract the distinguishing characteristics of brain tissue from tumors. The use of fusion images allows for more accurate detection of tumor types. In comparison with other approaches, the proposed method has demonstrated superior results. [ABSTRACT FROM AUTHOR]

Facial and object recognition are more and more applied in our life. Therefore, this field has become important to both academicians and practitioners. Face recognition systems are complex systems using features of the face to recognize. Current face recognition systems may be used to increase work efficiency in various methods, including smart homes, online banking, traffic, sports, robots, and others. With various applications like this, the number of facial recognition methods has been increasing in recent years. However, the performance of face recognition systems can be significantly affected by various factors such as lighting conditions, and different types of masks (sunglasses, scarves, hats, etc.). In this paper, a detailed comparison between face recognition techniques is exposed by listing the structure of each model, the advantages and disadvantages as well as performing experiments to demonstrate the robustness, accuracy, and complexity of each algorithm. To be detailed, let's give a performance comparison of three methods for measuring the efficacy of face recognition systems including a support vector machine (SVM), a visual geometry group with 16 layers (VGG-16), and a residual network with 50 layers (ResNet-50) in real-life settings. The efficiency of algorithms is evaluated in various environments such as normal light indoors, backlit indoors, low light indoors, natural light outdoors, and backlit outdoors. In addition, this paper also evaluates faces with hats and glasses to examine the accuracy of the methods. The experimental results indicate that the ResNet-50 has the highest accuracy to identify faces. The time to recognize is ranging from 1.1 s to 1.2 s in the normal environment. [ABSTRACT FROM AUTHOR]

In recent years, deep learning-based hashing techniques have garnered significant attention within the academic community, owing to their impressive performance achievements. However, the current deep hashing method is limited by its use of neural networks cannot make good use of the important feature information of the image, thus limiting the retrieval accuracy. In this paper, we design a two-channel feature extraction network called the local and global feature fusion network (LGDH). Specifically, the features extracted by visual geometry group (VGG) network and con-volution-enhanced image transformer (CeiT) network are deeply orthogonal fusion of local and global features, which solves the problem of lack of semantic correlation in existing deep hash methods. To fully harness the insights gained from previous hash function learning, we employ long short-term memory (LSTM) decoder to proficiently shape and create a resilient public binary code space. Through detailed experiments, our proposed method has achieved the Sota level in three public datasets CIFAR10, NUS-WIDE and MIRFLICKR compared with multiple cutting-edge algorithms, which can prove the effectiveness of our pro-posed method.

The proposal of perceptual loss solves the over-smoothing problem of images caused by pixel-wise loss and improves the visual quality of the images, but it also inevitably produces a large number of artifacts and distortions in images. The reason for this phenomenon is that the perceptual features only rely on a single pre-trained visual geometry group (VGG) network, which results in the features of the image being unable to be fully extracted, thus limiting the reasoning ability of the model. To fundamentally reduce the generation of artifacts and distortions, this paper proposes the Dual Perceptual Loss (DP Loss). First, we improve the perceptual feature extraction method so that it no longer only extracts single-type VGG features. In addition, a residual network (ResNet) feature that has a complementary relationship with the VGG feature can also be extracted. Then, we propose a dynamic weighting method to eliminate the magnitude difference between perceptual losses. Finally, to obtain the excellent effect of image reconstruction, enhanced super-resolution generative adversarial network (ESRGAN) with strong learning capability is used in this paper to adapt the complexity of DP Loss. The abundant experimental studies and evaluations are conducted on benchmark datasets. Results are encouraging and better than those previously reported on these datasets. The code is available at https://github.com/Sunny6-6-6/ESRGAN-DP.

Image Captioning (IC) is one of the most widely discussed topic in Artificial Intelligence. In this paper, Myanmar image caption is generated using EfficientNetB7 and Bidirectional Long Short-Term Memory (Bi-LSTM) with GloVe embedding and described about the comparative analysis results. For the purpose of achieving better performance, Myanmar image caption corpus is created and annotated over 50k sentences for 10k images, which are based on Flickr8k dataset and 2k images are selected from Flickr30k dataset. Two different types of segmentations such as word and syllable segmentation level are studied in text pre-processing step and then constructed our own GloVe vectors for both segmentations. As far as being aware and up to our knowledge, this is the first attempt of applying syllable and word vector features in neural network-based Myanmar IC system and then compared with one-hot encoding vectors on various different models. According to the evaluations results, EfficientNetB7 with Bi-LSTM using word and syllable GloVe embedding outperforms than EfficientNetB7 and Bi-LSTM with one-hot encoding, other neural networks such as Gated Recurrent Unit (GRU), Bidirectional Gated Recurrent Unit (Bi-GRU), and Long Short-Term Memory (LSTM), VGG16 with Bi-LSTM, NASNetLarge with Bi-LSTM models as well as baseline models. EffecientNetB7 with Bi-LSTM using GloVe vectors achieved the highest BLEU-4 score of 35.09%, 49.52% of ROUGE-L, 54.34% of ROUGE-SU4 and 21.3% of METEOR score on word vectors, and the highest BLEU-4 score of 46.2%, 65.62% of ROUGE-L, 68.43% of ROUGE-SU4 and 27.07% of METEOR score on syllable vectors. [ABSTRACT FROM AUTHOR]

Pneumonia is an acute respiratory infection that has led to significant deaths of people worldwide. This lung disease is more common in people older than 65 and children under five years old. Although the treatment of pneumonia can be challenging, it can be prevented by early diagnosis using Computer-Aided Diagnosis (CAD) systems. Chest X-Rays (CXRs) are currently the primary imaging tool for detection of pneumonia, which are widely used by radiologists. While the standard approach of detecting pneumonia is based on clinicians’ decisions, various Deep Learning (DL) methods have been developed for detection of pneumonia considering CAD system. In this regard, a novel hybrid Convolutional Neural Network (CNN) model is proposed using three classification approaches. In the first classification approach, Fully-Connected (FC) layers are utilized for the classification of CXR images. This model is trained for several epochs and the weights that result in the highest classification accuracy are saved. In the second classification approach, the trained optimized weights are utilized to extract the most representative CXR image features and Machine Learning (ML) classifiers are employed to classify the images. In the third classification approach, an ensemble of the proposed classifiers is created to classify CXR images. The results suggest that the proposed ensemble classifier using Support Vector Machine (SVM) with Radial Basis Function (RBF) and Logistic Regression (LR) classifiers has the best performance with 98.55% accuracy. Ultimately, this model is deployed to create a web-based CAD system to assist radiologists in pneumonia detection with a significant accuracy.

Emotions play a large role in interpersonal communication, marketing, healthcare and the service industry. For this reason, much research has been carried out on emotion classification until today. Audio-visual emotion recognition is a field within artificial intelligence and machine learning that focuses on recognizing and understanding human emotions from both visual and audio cues. It combines computer vision and audio processing techniques to analyze and interpret emotional states expressed by individuals. This paper presents a deep learning model developed over an operational neural network using multiple inputs and aimed at audio-visual emotion recognition. The proposed network utilizes both visual and audio information in an end to end approach. The primary objective of this work is to demonstrate that multi-input models can produce more efficient outcomes compared to single-input models in emotion classification. Another objective is to demonstrate the superior performance of weight calculation methods employed in operational neural networks compared to the conventional weight calculation methods used in convolutional neural networks. Therefore, we want to demonstrate that substituting convolutional neural network approaches with operational neural network methods can yield superior outcomes in emotion categorization models. In the proposed architecture, regular convolutional layers are replaced with operational layers. The experimental results demonstrate that the operational convolutional architecture performs better compared to the classical convolutional neural network architecture. [ABSTRACT FROM AUTHOR]

Novel corona virus COVID‐19 has spread rapidly all over the world. Due to increasing COVID‐19 cases, there is a dearth of testing kits. Therefore, there is a severe need for an automatic recognition system as a solution to reduce the spreading of the COVID‐19 virus. This work offers a decision support system based on the X‐ray image to diagnose the presence of the COVID‐19 virus. A deep learning‐based computer‐aided decision support system will be capable to differentiate between COVID‐19 and pneumonia. Recently, convolutional neural network (CNN) is designed for the diagnosis of COVID‐19 patients through chest radiography (or chest X‐ray, CXR) images. However, due to the usage of CNN, there are some limitations with these decision support systems. These systems suffer with the problem of view‐invariance and loss of information due to down‐sampling. In this paper, the capsule network (CapsNet)‐based system named visual geometry group capsule network (VGG‐CapsNet) for the diagnosis of COVID‐19 is proposed. Due to the usage of capsule network (CapsNet), the authors have succeeded in removing the drawbacks found in the CNN‐based decision support system for the detection of COVID‐19. Through simulation results, it is found that VGG‐CapsNet has performed better than the CNN‐CapsNet model for the diagnosis of COVID‐19. The proposed VGG‐CapsNet‐based system has shown 97% accuracy for COVID‐19 versus non‐COVID‐19 classification, and 92% accuracy for COVID‐19 versus normal versus viral pneumonia classification. Proposed VGG‐CapsNet‐based system available at https://github.com/shamiktiwari/COVID19%5fXray can be used to detect the existence of COVID‐19 virus in the human body through chest radiographic images. [ABSTRACT FROM AUTHOR]

The rapid growth and advancement in the application of remote sensing (RS) and geographic information systems (GIS) in several application domains helped many researchers to analyze a wide range of information from satellite imagery. Settlement information that includes building foot prints is a very essential parameter for different applications such as urban planning, environmental planning and disaster management. VGG-16 (Visual Geometry Group) conventional neural network (CNN) model is a popular model used to detect and classify the input images. In this study, a new approach was proposed by integration of VGG-16-CNN model with spectral and textural information of satellite images for identification of building foot prints. The model was trained and implemented to identify the building footprints using Worldview-3 high-resolution satellite image over part of Mumbai city of Maharashtra state, India. Classification accuracy in the proposed approach is observed to be nearly 94% as compared to 82% in case of single-shot detector (SSD) algorithm alone. Metric parameters such as F1 score of 0.957, intersection over union (IoU) of 94.86% and total error rate of 8.133% also indicated better performance of the proposed approach. Particularly, the approach is highly beneficial for urban development authorities as they need to monitor the large number of vacant lands spread across urban areas. [ABSTRACT FROM AUTHOR]

Optical character recognition (OCR) has become one of the most important techniques in computer vision, given that it can easily obtain digital information from various images on the Internet of Things (IoT). However, existing OCR techniques pose a big challenge in the recognition of the Chinese uppercase characters due to their poor performance. In order to solve the problem, this paper proposes a deep learning-aided OCR technique for improving recognition accuracy. First, we generate a database of the Chinese uppercase characters to train four neural networks: a convolution neural network (CNN), a visual geometry group, a capsule network, and a residual network. Second, the four networks are tested on the generated dataset in terms of accuracy, network weight, and test time. Finally, in order to reduce test time and save computational resources, we also develop a lightweight CNN method to prune the network weight by 96.5% while reducing accuracy by no more than 1.26%.

Copyright of Telecommunication Engineering is the property of Telecommunication Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Traffic density estimation is a very important component of an automated traffic monitoring system. Traffic density estimation can be used in a number of traffic applications – from congestion identification to macroscopic traffic control in urban environment. In this paper, we implemented Single Shot Detection (SSD) and MobileNet-SSD to estimate traffic density. SSD is capable of handling different shape, size and view angle of the objects. MobileNet-SSD is a cross-trained model from SSD to MobileNet architecture, which is faster than SSD. In this study, we show a key application area for the SSD and MobileNet-SSD framework. The advantages and shortcomings of the SSD and MobileNet-SSD framework were analyzed using fifty-nine individual traffic cameras. In addition, we compared the two algorithms with manually estimated density. The SSD framework shows significant potential in the field of traffic density estimation. SSD achieved 92.97% average detection accuracy in our experiment. On the other hand, the MobileNet-SSD achieved 79.30% average detection accuracy. • Real time traffic density estimation by traffic cameras and deep learning algorithms. • Car counting using SSD and MobileNet-SSD. • Quantitative analysis of the algorithms based on number of training data vs accuracy. • Performance and speed comparison of SSD and MobileNet-SSD. [ABSTRACT FROM AUTHOR]

Highlights • Computer-assisted automated classification of calcaneus fracture in CT images. • ResNet outperformed VGG in bone fracture classification. • High accuracy and reduced runtimes achieved in automatic fracture detection using SURF algorithm. • Demonstrate feasibility of using deep learning neural network in automatic calcaneal fracture detection. Abstract Background and objectives The calcaneus is the most fracture-prone tarsal bone and injuries to the surrounding tissue are some of the most difficult to treat. Currently there is a lack of consensus on treatment or interpretation of computed tomography (CT) images for calcaneus fractures. This study proposes a novel computer-assisted method for automated classification and detection of fracture locations in calcaneus CT images using a deep learning algorithm. Methods Two types of Convolutional Neural Network (CNN) architectures with different network depths, a Residual network (ResNet) and a Visual geometry group (VGG), were evaluated and compared for the classification performance of CT scans into fracture and non-fracture categories based on coronal, sagittal, and transverse views. The bone fracture detection algorithm incorporated fracture area matching using the speeded-up robust features (SURF) method, Canny edge detection, and contour tracing. Results Results showed that ResNet was comparable in accuracy (98%) to the VGG network for bone fracture classification but achieved better performance for involving a deeper neural network architecture. ResNet classification results were used as the input for detecting the location and type of bone fracture using SURF algorithm. Conclusions Results from real patient fracture data sets demonstrate the feasibility using deep CNN and SURF for computer-aided classification and detection of the location of calcaneus fractures in CT images. [ABSTRACT FROM AUTHOR]

Colorectal cancer (CRC) is a major global health concern. Its early diagnosis is extremely important, as it determines treatment options and strongly influences the length of survival. Histologic diagnosis can be made by pathologists based on images of tissues obtained from a colonoscopic biopsy. Convolutional neural networks (CNNs)—i.e., deep neural networks (DNNs) specifically adapted to image data—have been employed to effectively classify or locate tumors in many types of cancer. Colorectal histology images of 28 normal and 29 tumor samples were obtained from the National Cancer Center, South Korea, and cropped into 6806 normal and 3474 tumor images. We developed five modifications of the system from the Visual Geometry Group (VGG), the winning entry in the classification task in the 2014 ImageNet Large Scale Visual Recognition Competition (ILSVRC) and examined them in two experiments. In the first experiment, we determined the best modified VGG configuration for our partial dataset, resulting in accuracies of 82.50%, 87.50%, 87.50%, 91.40%, and 94.30%, respectively. In the second experiment, the best modified VGG configuration was applied to evaluate the performance of the CNN model. Subsequently, using the entire dataset on the modified VGG-E configuration, the highest results for accuracy, loss, sensitivity, and specificity, respectively, were 93.48%, 0.4385, 95.10%, and 92.76%, which equates to correctly classifying 294 normal images out of 309 and 667 tumor images out of 719. [ABSTRACT FROM AUTHOR]

The proposal of perceptual loss solves the over-smoothing problem of images caused by pixel-wise loss and improves the visual quality of the images, but it also inevitably produces a large number of artifacts and distortions in images. The reason for this phenomenon is that the perceptual features only rely on a single pre-trained visual geometry group (VGG) network, which results in the features of the image being unable to be fully extracted, thus limiting the reasoning ability of the model. To fundamentally reduce the generation of artifacts and distortions, this paper proposes the Dual Perceptual Loss (DP Loss). First, we improve the perceptual feature extraction method so that it no longer only extracts single-type VGG features. In addition, a residual network (ResNet) feature that has a complementary relationship with the VGG feature can also be extracted. Then, we propose a dynamic weighting method to eliminate the magnitude difference between perceptual losses. Finally, to obtain the excellent effect of image reconstruction, enhanced super-resolution generative adversarial network (ESRGAN) with strong learning capability is used in this paper to adapt the complexity of DP Loss. The abundant experimental studies and evaluations are conducted on benchmark datasets. Results are encouraging and better than those previously reported on these datasets. The code is available at https://github.com/Sunny6-6-6/ESRGAN-DP., Competing Interests: The authors declare no conflict of interest., (© 2023 The Authors.)

Background: Ultrasound (US) is widely used in the clinical diagnosis of thyroid nodules. Artificial intelligence-powered US is becoming an important issue in the research community. This study aimed to develop an improved deep learning model-based algorithm to classify benign and malignant thyroid nodules (TNs) using thyroid US images., Methods: In total, 592 patients with 600 TNs were included in the internal training, validation, and testing data set; 187 patients with 200 TNs were recruited for the external test data set. We developed a Visual Geometry Group (VGG)-16T model, based on the VGG-16 architecture, but with additional batch normalization (BN) and dropout layers in addition to the fully connected layers. We conducted a 10-fold cross-validation to analyze the performance of the VGG-16T model using a data set of gray-scale US images from 5 different brands of US machines., Results: For the internal data set, the VGG-16T model had 87.43% sensitivity, 85.43% specificity, and 86.43% accuracy. For the external data set, the VGG-16T model achieved an area under the curve (AUC) of 0.829 [95% confidence interval (CI): 0.770-0.879], a radiologist with 15 years' working experience achieved an AUC of 0.705 (95% CI: 0.659-0.801), a radiologist with 10 years' experience achieved an AUC of 0.725 (95% CI: 0.653-0.797), and a radiologist with 5 years' experience achieved an AUC of 0.660 (95% CI: 0.584-0.736)., Conclusions: The VGG-16T model had high specificity, sensitivity, and accuracy in differentiating between malignant and benign TNs. Its diagnostic performance was superior to that of experienced radiologists. Thus, the proposed improved deep-learning model can assist radiologists to diagnose thyroid cancer., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://dx.doi.org/10.21037/atm-21-4328). The authors have no conflicts of interest to declare., (2021 Annals of Translational Medicine. All rights reserved.)