MDPI - Publisher of Open Access Journals

44 pages, 1546 KiB

Open AccessReview

Deep Learning for 3D Reconstruction, Augmentation, and Registration: A Review Paper

by Prasoon Kumar Vinodkumar, Dogus Karabulut, Egils Avots, Cagri Ozcinar and Gholamreza Anbarjafari

Entropy 2024, 26(3), 235; https://doi.org/10.3390/e26030235 - 7 Mar 2024

Cited by 10 | Viewed by 10089

The research groups in computer vision, graphics, and machine learning have dedicated a substantial amount of attention to the areas of 3D object reconstruction, augmentation, and registration. Deep learning is the predominant method used in artificial intelligence for addressing computer vision challenges. However, [...] Read more.

The research groups in computer vision, graphics, and machine learning have dedicated a substantial amount of attention to the areas of 3D object reconstruction, augmentation, and registration. Deep learning is the predominant method used in artificial intelligence for addressing computer vision challenges. However, deep learning on three-dimensional data presents distinct obstacles and is now in its nascent phase. There have been significant advancements in deep learning specifically for three-dimensional data, offering a range of ways to address these issues. This study offers a comprehensive examination of the latest advancements in deep learning methodologies. We examine many benchmark models for the tasks of 3D object registration, augmentation, and reconstruction. We thoroughly analyse their architectures, advantages, and constraints. In summary, this report provides a comprehensive overview of recent advancements in three-dimensional deep learning and highlights unresolved research areas that will need to be addressed in the future. Full article

(This article belongs to the Special Issue Entropy in Machine Learning Applications)

► Show Figures

Figure 1

35 pages, 2329 KiB

Open AccessReview

A Survey on Deep Learning Based Segmentation, Detection and Classification for 3D Point Clouds

by Prasoon Kumar Vinodkumar, Dogus Karabulut, Egils Avots, Cagri Ozcinar and Gholamreza Anbarjafari

Entropy 2023, 25(4), 635; https://doi.org/10.3390/e25040635 - 10 Apr 2023

Cited by 27 | Viewed by 9270

Abstract

The computer vision, graphics, and machine learning research groups have given a significant amount of focus to 3D object recognition (segmentation, detection, and classification). Deep learning approaches have lately emerged as the preferred method for 3D segmentation problems as a result of their [...] Read more.

The computer vision, graphics, and machine learning research groups have given a significant amount of focus to 3D object recognition (segmentation, detection, and classification). Deep learning approaches have lately emerged as the preferred method for 3D segmentation problems as a result of their outstanding performance in 2D computer vision. As a result, many innovative approaches have been proposed and validated on multiple benchmark datasets. This study offers an in-depth assessment of the latest developments in deep learning-based 3D object recognition. We discuss the most well-known 3D object recognition models, along with evaluations of their distinctive qualities. Full article

(This article belongs to the Topic Machine and Deep Learning)

► Show Figures

Figure 1

20 pages, 32917 KiB

Open AccessArticle

Towards Automated Detection and Localization of Red Deer Cervus elaphus Using Passive Acoustic Sensors during the Rut

by Egils Avots, Alekss Vecvanags, Jevgenijs Filipovs, Agris Brauns, Gundars Skudrins, Gundega Done, Janis Ozolins, Gholamreza Anbarjafari and Dainis Jakovels

Remote Sens. 2022, 14(10), 2464; https://doi.org/10.3390/rs14102464 - 20 May 2022

Cited by 4 | Viewed by 3182

Abstract

Passive acoustic sensors have the potential to become a valuable complementary component in red deer Cervus elaphus monitoring providing deeper insight into the behavior of stags during the rutting period. Automation of data acquisition and processing is crucial for adaptation and wider uptake [...] Read more.

Passive acoustic sensors have the potential to become a valuable complementary component in red deer Cervus elaphus monitoring providing deeper insight into the behavior of stags during the rutting period. Automation of data acquisition and processing is crucial for adaptation and wider uptake of acoustic monitoring. Therefore, an automated data processing workflow concept for red deer call detection and localization was proposed and demonstrated. The unique dataset of red deer calls during the rut in September 2021 was collected with four GPS time-synchronized microphones. Five supervised machine learning algorithms were tested and compared for the detection of red deer rutting calls where the support-vector-machine-based approach demonstrated the best performance of −96.46% detection accuracy. For sound source location, a hyperbolic localization approach was applied. A novel approach based on cross-correlation and spectral feature similarity was proposed for sound delay assessment in multiple microphones resulting in the median localization error of 16 m, thus providing a solution for automated sound source localization—the main challenge in the automation of the data processing workflow. The automated approach outperformed manual sound delay assessment by a human expert where the median localization error was 43 m. Artificial sound records with a known location in the pilot territory were used for localization performance testing. Full article

(This article belongs to the Special Issue Remote Sensing for Applied Wildlife Ecology)

► Show Figures

Graphical abstract

13 pages, 8371 KiB

Open AccessArticle

Ungulate Detection and Species Classification from Camera Trap Images Using RetinaNet and Faster R-CNN

by Alekss Vecvanags, Kadir Aktas, Ilja Pavlovs, Egils Avots, Jevgenijs Filipovs, Agris Brauns, Gundega Done, Dainis Jakovels and Gholamreza Anbarjafari

Entropy 2022, 24(3), 353; https://doi.org/10.3390/e24030353 - 28 Feb 2022

Cited by 48 | Viewed by 5217

Abstract

Changes in the ungulate population density in the wild has impacts on both the wildlife and human society. In order to control the ungulate population movement, monitoring systems such as camera trap networks have been implemented in a non-invasive setup. However, such systems [...] Read more.

Changes in the ungulate population density in the wild has impacts on both the wildlife and human society. In order to control the ungulate population movement, monitoring systems such as camera trap networks have been implemented in a non-invasive setup. However, such systems produce a large number of images as the output, hence making it very resource consuming to manually detect the animals. In this paper, we present a new dataset of wild ungulates which was collected in Latvia. Moreover, we demonstrate two methods, which use RetinaNet and Faster R-CNN as backbones, respectively, to detect the animals in the images. We discuss the optimization of training and impact of data augmentation on the performance. Finally, we show the result of aforementioned tune networks over the real world data collected in Latvia. Full article

(This article belongs to the Section Multidisciplinary Applications)

► Show Figures

Figure 1

14 pages, 390 KiB

Open AccessArticle

Ensemble Approach for Detection of Depression Using EEG Features

by Egils Avots, Klāvs Jermakovs, Maie Bachmann, Laura Päeske, Cagri Ozcinar and Gholamreza Anbarjafari

Entropy 2022, 24(2), 211; https://doi.org/10.3390/e24020211 - 28 Jan 2022

Cited by 46 | Viewed by 4635

Abstract

Depression is a public health issue that severely affects one’s well being and can cause negative social and economic effects to society. To raise awareness of these problems, this research aims at determining whether the long-lasting effects of depression can be determined from [...] Read more.

Depression is a public health issue that severely affects one’s well being and can cause negative social and economic effects to society. To raise awareness of these problems, this research aims at determining whether the long-lasting effects of depression can be determined from electroencephalographic (EEG) signals. The article contains an accuracy comparison for SVM, LDA, NB, kNN, and D3 binary classifiers, which were trained using linear (relative band power, alpha power variability, spectral asymmetry index) and nonlinear (Higuchi fractal dimension, Lempel–Ziv complexity, detrended fluctuation analysis) EEG features. The age- and gender-matched dataset consisted of 10 healthy subjects and 10 subjects diagnosed with depression at some point in their lifetime. Most of the proposed feature selection and classifier combinations achieved accuracy in the range of 80% to 95%, and all the models were evaluated using a 10-fold cross-validation. The results showed that the motioned EEG features used in classifying ongoing depression also work for classifying the long-lasting effects of depression. Full article

► Show Figures

Figure 1

15 pages, 9107 KiB

Open AccessArticle

Privacy-Constrained Biometric System for Non-Cooperative Users

by Mohammad N. S. Jahromi, Pau Buch-Cardona, Egils Avots, Kamal Nasrollahi, Sergio Escalera, Thomas B. Moeslund and Gholamreza Anbarjafari

Entropy 2019, 21(11), 1033; https://doi.org/10.3390/e21111033 - 24 Oct 2019

Cited by 14 | Viewed by 3776

Abstract

With the consolidation of the new data protection regulation paradigm for each individual within the European Union (EU), major biometric technologies are now confronted with many concerns related to user privacy in biometric deployments. When individual biometrics are disclosed, the sensitive information about [...] Read more.

With the consolidation of the new data protection regulation paradigm for each individual within the European Union (EU), major biometric technologies are now confronted with many concerns related to user privacy in biometric deployments. When individual biometrics are disclosed, the sensitive information about his/her personal data such as financial or health are at high risk of being misused or compromised. This issue can be escalated considerably over scenarios of non-cooperative users, such as elderly people residing in care homes, with their inability to interact conveniently and securely with the biometric system. The primary goal of this study is to design a novel database to investigate the problem of automatic people recognition under privacy constraints. To do so, the collected data-set contains the subject’s hand and foot traits and excludes the face biometrics of individuals in order to protect their privacy. We carried out extensive simulations using different baseline methods, including deep learning. Simulation results show that, with the spatial features extracted from the subject sequence in both individual hand or foot videos, state-of-the-art deep models provide promising recognition performance. Full article

(This article belongs to the Special Issue Statistical Machine Learning for Human Behaviour Analysis)

► Show Figures

Figure 1

20 pages, 8836 KiB

Open AccessArticle

Virtual Reality and Its Applications in Education: Survey

by Dorota Kamińska, Tomasz Sapiński, Sławomir Wiak, Toomas Tikk, Rain Eric Haamer, Egils Avots, Ahmed Helmi, Cagri Ozcinar and Gholamreza Anbarjafari

Information 2019, 10(10), 318; https://doi.org/10.3390/info10100318 - 16 Oct 2019

Cited by 332 | Viewed by 61681

Abstract

In the education process, students face problems with understanding due to the complexity, necessity of abstract thinking and concepts. More and more educational centres around the world have started to introduce powerful new technology-based tools that help meet the needs of the diverse [...] Read more.

In the education process, students face problems with understanding due to the complexity, necessity of abstract thinking and concepts. More and more educational centres around the world have started to introduce powerful new technology-based tools that help meet the needs of the diverse student population. Over the last several years, virtual reality (VR) has moved from being the purview of gaming to professional development. It plays an important role in teaching process, providing an interesting and engaging way of acquiring information. What follows is an overview of the big trend, opportunities and concerns associated with VR in education. We present new opportunities in VR and put together the most interesting, recent virtual reality applications used in education in relation to several education areas such as general, engineering and health-related education. Additionally, this survey contributes by presenting methods for creating scenarios and different approaches for testing and validation. Lastly, we conclude and discuss future directions of VR and its potential to improve the learning experience. Full article

(This article belongs to the Section Information and Communications Technology)

► Show Figures

Figure 1

19 pages, 1317 KiB

Open AccessArticle

Action Recognition Using Single-Pixel Time-of-Flight Detection

by Ikechukwu Ofodile, Ahmed Helmi, Albert Clapés, Egils Avots, Kerttu Maria Peensoo, Sandhra-Mirella Valdma, Andreas Valdmann, Heli Valtna-Lukner, Sergey Omelkov, Sergio Escalera, Cagri Ozcinar and Gholamreza Anbarjafari

Entropy 2019, 21(4), 414; https://doi.org/10.3390/e21040414 - 18 Apr 2019

Cited by 9 | Viewed by 5509

Abstract

Action recognition is a challenging task that plays an important role in many robotic systems, which highly depend on visual input feeds. However, due to privacy concerns, it is important to find a method which can recognise actions without using visual feed. In [...] Read more.

Action recognition is a challenging task that plays an important role in many robotic systems, which highly depend on visual input feeds. However, due to privacy concerns, it is important to find a method which can recognise actions without using visual feed. In this paper, we propose a concept for detecting actions while preserving the test subject’s privacy. Our proposed method relies only on recording the temporal evolution of light pulses scattered back from the scene. Such data trace to record one action contains a sequence of one-dimensional arrays of voltage values acquired by a single-pixel detector at 1 GHz repetition rate. Information about both the distance to the object and its shape are embedded in the traces. We apply machine learning in the form of recurrent neural networks for data analysis and demonstrate successful action recognition. The experimental results show that our proposed method could achieve on average

96.47 %

accuracy on the actions walking forward, walking backwards, sitting down, standing up and waving hand, using recurrent neural network. Full article

(This article belongs to the Special Issue Statistical Machine Learning for Human Behaviour Analysis)

► Show Figures

Figure 1

Search Results (8)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Saved Queries

Search Filter Reset All

Years

Feature Papers

Subjects

Journals

Article Types

Countries / Regions

Search Results (8)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI