scholarly journals A Novel Method for Goal Recognition from 10 m Distance Using Deep Learning in CanSat

2021 ◽  
Vol 33 (6) ◽  
pp. 1359-1372
Author(s):  
Miho Akiyama ◽  
Takuya Saito ◽  
◽  
Keyword(s):  
Deep Learning ◽  
Image Classification ◽  
Recognition Rate ◽  
Computation Time ◽  
Regions Of Interest ◽  
Effective Number ◽  
Novel Method ◽  
Goal Recognition ◽  
Almost All ◽  
First Time

In this study, we propose a method for CanSat to recognize and guide a goal using deep learning image classification even 10 m away from the goal, and describe the results of demonstrative evaluation to confirm the effectiveness of the method. We applied deep learning image classification to goal recognition in CanSat for the first time at ARLISS 2019, and succeeded in guiding it almost all the way to the goal in all three races, winning the first place as overall winner. However, the conventional method has a drawback in that the goal recognition rate drops significantly when the CanSat is more than 6–7 m away from the goal, making it difficult to guide the CanSat to the goal when it moves away from the goal because of various factors. To enable goal recognition from a distance of 10 m from the goal, we investigated the number of horizontal regions of interest divisions and the method of vertical shifts during image recognition, and clarified the effective number of divisions and recognition rate using experiments. Although object detection is commonly used to detect the position of an object from an image by deep learning, we confirmed that the proposed method has a higher recognition rate at long distances and a shorter computation time than SSD MobileNet V1. In addition, we participated in the CanSat contest ACTS 2020 to evaluate the effectiveness of the proposed method and achieved the zero-distance goal in all three competitions, demonstrating its effectiveness by winning first place in the comeback category.

scholarly journals Shelf Auditing Based on Image Classification Using Semi-Supervised Deep Learning to Increase On-Shelf Availability in Grocery Stores

Sensors ◽  
2021 ◽  
Vol 21 (2) ◽  
pp. 327
Author(s):  
Ramiz Yilmazer ◽  
Derya Birant
Keyword(s):  
Artificial Intelligence ◽  
Computer Vision ◽  
Deep Learning ◽  
Image Classification ◽  
Experimental Studies ◽  
Grocery Stores ◽  
Software Application ◽  
Explainable Artificial Intelligence ◽  
Key Factor ◽  
First Time

Providing high on-shelf availability (OSA) is a key factor to increase profits in grocery stores. Recently, there has been growing interest in computer vision approaches to monitor OSA. However, the largest and well-known computer vision datasets do not provide annotation for store products, and therefore, a huge effort is needed to manually label products on images. To tackle the annotation problem, this paper proposes a new method that combines two concepts “semi-supervised learning” and “on-shelf availability” (SOSA) for the first time. Moreover, it is the first time that “You Only Look Once” (YOLOv4) deep learning architecture is used to monitor OSA. Furthermore, this paper provides the first demonstration of explainable artificial intelligence (XAI) on OSA. It presents a new software application, called SOSA XAI, with its capabilities and advantages. In the experimental studies, the effectiveness of the proposed SOSA method was verified on image datasets, with different ratios of labeled samples varying from 20% to 80%. The experimental results show that the proposed approach outperforms the existing approaches (RetinaNet and YOLOv3) in terms of accuracy.

COVID-19 improved diagnoses based on the open-morphology filter and deep-learning

2021 ◽  
Vol 15 ◽  
Author(s):  
Majid Dherar Younus ◽  
Mohammad J M Zedan ◽  
Fahad Layth Malallah ◽  
Mustafa Ghanem Saeed
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Detection Method ◽  
Recognition Rate ◽  
Processing Technique ◽  
Ct Images ◽  
Ct Scans ◽  
Computer Aided ◽  
First Time ◽  
Morphology Filter

Background: Coronavirus (COVID-19) has appeared first time in Wuhan, China, as an acute respiratory syndrome and spread rapidly. It has been declared a pandemic by the WHO. Thus, there is an urgent need to develop an accurate computer-aided method to assist clinicians in identifying COVID-19-infected patients by computed tomography CT images. The contribution of this paper is that it proposes a pre-processing technique that increases the recognition rate compared to the techniques existing in the literature. Methods: The proposed pre-processing technique, which consists of both contrast enhancement and open-morphology filter, is highly effective in decreasing the diagnosis error rate. After carrying out pre-processing, CT images are fed to a 15-layer convolution neural network (CNN) as deep-learning for the training and testing operations. The dataset used in this research has been publically published, in which CT images were collected from hospitals in Sao Paulo, Brazil. This dataset is composed of 2482 CT scans images, which include 1252 CT scans of SARS-CoV-2 infected patients and 1230 CT scans of non-infected SARS-CoV-2 patients. Results: The proposed detection method achieves up to 97.8% accuracy, which outperforms the reported accuracy of the dataset by 97.3%. Conclusion: The performance in terms of accuracy has been improved up to 0.5% by the proposed methodology over the published dataset and its method.

scholarly journals A new method based on stacked auto-encoders to identify abnormal weather radar echo images

2020 ◽  
Vol 2020 (1) ◽  
Author(s):  
Ling Yang ◽  
Yun Wang ◽  
Zhongke Wang ◽  
Yang Qi ◽  
Yong Li ◽  
...  
Keyword(s):  
Deep Learning ◽  
Recognition Rate ◽  
Weather Radar ◽  
Practical Significance ◽  
Automatic Identification ◽  
Hardware Failure ◽  
Radar Echoes ◽  
Radar Echo ◽  
Novel Method ◽  
Classical Image

Abstract It is not denied that real-time monitoring of radar products is an important part in actual meteorological operations. But the weather radar often brings out abnormal radar echoes due to various factors, such as climate and hardware failure. So it is of great practical significance and research value to realize automatic identification of radar anomaly products. However, the traditional algorithms to identify anomalies of weather radar echo images are not the most accurate and efficient. In order to improve the efficiency of the anomaly identification, a novel method combining the theory of classical image processing and deep learning was proposed. The proposed method mainly includes three parts: coordinate transformation, integral projection, and classification using deep learning. Furthermore, extensive experiments have been done to validate the performance of the new algorithm. The results show that the recognition rate of the proposed method can reach up to more than 95%, which can successfully achieve the goal of screening abnormal radar echo images; also, the computation speed of it is fairly satisfactory.

scholarly journals Pendeteksi Citra Masker Wajah Menggunakan CNN dan Transfer Learning

2021 ◽  
Vol 8 (6) ◽  
pp. 1293
Author(s):  
Mohammad Farid Naufal ◽  
Selvia Ferdiana Kusuma
Keyword(s):  
Neural Network ◽  
Computer Vision ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Image Classification ◽  
Transfer Learning ◽  
Cross Validation ◽  
Learning Algorithm ◽  
Computing Time ◽  
Computation Time

<p class="Abstrak">Pada tahun 2021 pandemi Covid-19 masih menjadi masalah di dunia. Protokol kesehatan diperlukan untuk mencegah penyebaran Covid-19. Penggunaan masker wajah adalah salah satu protokol kesehatan yang umum digunakan. Pengecekan secara manual untuk mendeteksi wajah yang tidak menggunakan masker adalah pekerjaan yang lama dan melelahkan. Computer vision merupakan salah satu cabang ilmu komputer yang dapat digunakan untuk klasifikasi citra. Convolutional Neural Network (CNN) merupakan algoritma deep learning yang memiliki performa bagus dalam klasifikasi citra. Transfer learning merupakan metode terkini untuk mempercepat waktu training pada CNN dan untuk mendapatkan performa klasifikasi yang lebih baik. Penelitian ini melakukan klasifikasi citra wajah untuk membedakan orang menggunakan masker atau tidak dengan menggunakan CNN dan Transfer Learning. Arsitektur CNN yang digunakan dalam penelitian ini adalah MobileNetV2, VGG16, DenseNet201, dan Xception. Berdasarkan hasil uji coba menggunakan 5-cross validation, Xception memiliki akurasi terbaik yaitu 0.988 dengan waktu total komputasi training dan testing sebesar 18274 detik. MobileNetV2 memiliki waktu total komputasi tercepat yaitu 4081 detik dengan akurasi sebesar 0.981.</p><p class="Abstrak"> </p><p class="Abstrak"><em><strong>Abstract</strong></em></p><p class="Judul2"><em>In 2021 the Covid-19 pandemic is still a problem in the world. Therefore, health protocols are needed to prevent the spread of Covid-19. The use of face masks is one of the commonly used health protocols. However, manually checking to detect faces that are not wearing masks is a long and tiring job. Computer vision is a branch of computer science that can be used for image classification. Convolutional Neural Network (CNN) is a deep learning algorithm that has good performance in image classification. Transfer learning is the latest method to speed up CNN training and get better classification performance. This study performs facial image classification to distinguish people using masks or not by using CNN and Transfer Learning. The CNN architecture used in this research is MobileNetV2, VGG16, DenseNet201, and Xception. Based on the results of trials using 5-cross validation, Xception has the best accuracy of 0.988 with a total computation time of training and testing of 18274 seconds. MobileNetV2 has the fastest total computing time of 4081 seconds with an accuracy of 0.981.</em></p><p class="Abstrak"><em><strong><br /></strong></em></p>

scholarly journals Particle Mobility Analysis Using Deep Learning and the Moment Scaling Spectrum

2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Marloes Arts ◽  
Ihor Smal ◽  
Maarten W. Paul ◽  
Claire Wyman ◽  
Erik Meijering
Keyword(s):  
Deep Learning ◽  
Time Lapse ◽  
Parameter Extraction ◽  
Mobility Analysis ◽  
Biologically Relevant ◽  
Particle Mobility ◽  
Wide Range ◽  
Novel Method ◽  
The Moment ◽  
First Time

AbstractQuantitative analysis of dynamic processes in living cells using time-lapse microscopy requires not only accurate tracking of every particle in the images, but also reliable extraction of biologically relevant parameters from the resulting trajectories. Whereas many methods exist to perform the tracking task, there is still a lack of robust solutions for subsequent parameter extraction and analysis. Here a novel method is presented to address this need. It uses for the first time a deep learning approach to segment single particle trajectories into consistent tracklets (trajectory segments that exhibit one type of motion) and then performs moment scaling spectrum analysis of the tracklets to estimate the number of mobility classes and their associated parameters, providing rich fundamental knowledge about the behavior of the particles under study. Experiments on in-house datasets as well as publicly available particle tracking data for a wide range of proteins with different dynamic behavior demonstrate the broad applicability of the method.

Deep Learning Based High-Resolution Remote Sensing Image classification

2017 ◽  
Vol 7 (10) ◽  
pp. 22
Author(s):  
Sumit Kaur
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Deep Learning ◽  
Image Classification ◽  
Language Processing ◽  
Object Perception ◽  
Remote Sensing Image ◽  
Research Area ◽  
Remote Sensing Image Classification ◽  
Unsupervised Algorithms

Abstract- Deep learning is an emerging research area in machine learning and pattern recognition field which has been presented with the goal of drawing Machine Learning nearer to one of its unique objectives, Artificial Intelligence. It tries to mimic the human brain, which is capable of processing and learning from the complex input data and solving different kinds of complicated tasks well. Deep learning (DL) basically based on a set of supervised and unsupervised algorithms that attempt to model higher level abstractions in data and make it self-learning for hierarchical representation for classification. In the recent years, it has attracted much attention due to its state-of-the-art performance in diverse areas like object perception, speech recognition, computer vision, collaborative filtering and natural language processing. This paper will present a survey on different deep learning techniques for remote sensing image classification. 

Organic Reactivity Control by Means of Neighboring Groups and Organometallics. A Personal Account

1994 ◽  
Vol 59 (1) ◽  
pp. 1-74 ◽  
Author(s):  
Pavel Kočovský
Keyword(s):  
Transition Metals ◽  
Cyclopropane Ring ◽  
Allylic Substitutions ◽  
Recent Developments ◽  
Cardioactive Drug ◽  
Novel Method ◽  
Metal Catalyzed ◽  
Electrophilic Additions ◽  
Near Future ◽  
First Time

This review summarizes the main topics of our research and covers the period of the last 15 years. The prime interest is focused on various ways of controlling the regio- and stereoselectivity of selected organic reactions, in particular electrophilic additions, cleavage of cyclopropane rings, and allylic substitutions by means of neighboring groups and/or transition and non-transition metals. In the first part, the factors governing the course of electrophilic additions are assessed, culminating in the formulation of selection rules for the reactivity of cyclohexene systems, and in a concise synthesis of the natural cardioactive drug, strophanthidin. These studies also contribute to a better understanding of the mechanisms of electrophilic additions. The second part describes recent developments in the stereo- and regiocontrolled cleavage of cyclopropane rings by non-transition metals (Tl and Hg), and the reactivity and transmetalation (with Pd) of the primary products. This methodology has resulted in novel routes to unique polycyclic structures, and will have synthetic applications in the near future. Evidence for the stereospecific "corner" cleavage of the cyclopropane ring has been provided for the first time for Tl and later for Hg. The third part deals with transition metal-catalyzed allylic substitution. Evidence for a new "syn" mechanism for the formation of the intermediate (π-allyl)palladium complex has been provided, which runs counter to the generally accepted "anti" mechanism. A novel method for a Pd-catalyzed allylic oxidation has been developed and employed in the synthesis of natural sesquiterpenes. The increasing importance of transition and non-transition metals for synthetic organic chemistry is demonstrated by their unique reactivity in a number of the papers included in this review.

scholarly journals Real UAV-Bird Image Classification Using CNN with a Synthetic Dataset

2021 ◽  
Vol 11 (9) ◽  
pp. 3863
Author(s):  
Ali Emre Öztürk ◽  
Ergun Erçelebi
Keyword(s):  
Deep Learning ◽  
Image Classification ◽  
Synthetic Data ◽  
Real Data ◽  
Corner Detection ◽  
Batch Size ◽  
Test Accuracy ◽  
Classification Problems ◽  
Auc Value ◽  
Classification Test

A large amount of training image data is required for solving image classification problems using deep learning (DL) networks. In this study, we aimed to train DL networks with synthetic images generated by using a game engine and determine the effects of the networks on performance when solving real-image classification problems. The study presents the results of using corner detection and nearest three-point selection (CDNTS) layers to classify bird and rotary-wing unmanned aerial vehicle (RW-UAV) images, provides a comprehensive comparison of two different experimental setups, and emphasizes the significant improvements in the performance in deep learning-based networks due to the inclusion of a CDNTS layer. Experiment 1 corresponds to training the commonly used deep learning-based networks with synthetic data and an image classification test on real data. Experiment 2 corresponds to training the CDNTS layer and commonly used deep learning-based networks with synthetic data and an image classification test on real data. In experiment 1, the best area under the curve (AUC) value for the image classification test accuracy was measured as 72%. In experiment 2, using the CDNTS layer, the AUC value for the image classification test accuracy was measured as 88.9%. A total of 432 different combinations of trainings were investigated in the experimental setups. The experiments were trained with various DL networks using four different optimizers by considering all combinations of batch size, learning rate, and dropout hyperparameters. The test accuracy AUC values for networks in experiment 1 ranged from 55% to 74%, whereas the test accuracy AUC values in experiment 2 networks with a CDNTS layer ranged from 76% to 89.9%. It was observed that the CDNTS layer has considerable effects on the image classification accuracy performance of deep learning-based networks. AUC, F-score, and test accuracy measures were used to validate the success of the networks.

Sign in / Sign up

Export Citation Format

Share Document