272 research outputs found
Multimodal Segmentation Based On A Novel 3d U-Net Deep Learning Architecture
No full textIn this paper, we propose a new approach for brain image segmentation based on a novel 3D U-Net deep fusion scheme. The proposed approach takes into consideration a fusion of multiple scan modalities including FLAIR, T1, T1Gd and T2, and by using a stacked CNN based 3D U-Net architecture allows modelling of multiclass segmentation of Gliomas, an aggressive form of brain tumours. The proposed model performs well for low resource settings, and requires lesser resource requirements, and with imbalanced class distribution, and natural data augmentation, by transforming 3D volumes to 2D sequences. An extensive quantitative and qualitative experimental evaluation of the proposed model in terms of dice score and dice loss performance metrics, for two publicly available datasets, corresponding to 2018 BraTS and 2021 BraTS challenge segmentation task, shows improved performance and generalization capability of the proposed lightweight model. </p
Multimodal Fusion for Robust Identity Authentication: Role of Liveness Checks
Full text linkMost of the current biometric identity authentication systems currently deployed are based on modeling the identity of a person based on unimodal information, i.e. face, voice, or fingerprint features. Also, many current interactive civilian remote human computer interaction applications are based on speech based voice features, which achieve significantly lower performance for operating environments with low signal-to-noise ratios (SNR). For a long time, use of acoustic information alone has been a great success for several automatic speech processing applications such as automatic speech transcription or speaker authentication, while face identification systems based visual information alone from faces also proved to be of equally successful. However, in adverse operating environments, performance of either of these systems could be suboptimal. Use of both visual and audio information can lead to better robustness, as they can provide complementary secondary clues that can help in the analysis of the primary biometric signals (Potamianos et al (2004)). The joint analysis of acoustic and visual speech can improve the robustness of automatic speech recognition systems (Liu et al (2002), Gurbuz et al (2002
An Automatic Decision Support System for Assessing SDG Contributions
No full textIn this paper, we propose an innovative computer-based decision support scheme based on Artificial Intelligence, consisting of novel text mining and machine learning techniques, to assess an organizational commitment to sustainable practices, with a granular analysis of the text content in their documents, and evaluate their conformance and alignment to one or more of the UN’s sustainable development goals. The proposed decision support system can help assist an organization to self-assess their business practices and marketing messages, in terms of presence in the social media channels, company documents, and refine their corporate vision and responsibility statements appropriately.</p
Deep Learning Based Decision Support Framework for Cardiovascular Disease Prediction
No full textIn this paper we propose a novel decision support framework based on deep learning for cardiovascular disease prediction. The proposed framework based on an innovative stacked dense neural layer and convolution neural network cascade architecture, addresses the significant imbalance in class distribution in CVD event detection task. The experimental evaluation of the proposed model was done on the NHANES super-dataset, obtained by fusion of different subsets of publicly NHANES (National Health and Nutrition Examination Survey) data for prediction of cardiovascular disease. Many machines and deep learning models have been proposed in the literature for CVD event detection. However, they assume balanced class distribution between positive and negative disease classes. For clinical settings, there is significant class imbalance, with few positive class samples as compared to abundant samples from normal or control class. Hence most of the traditional machine and deep learning models are vulnerable to class imbalance, even after using class-specific adjustment of weights (well established method for handling class imbalance) and can lead to poor performance for the minority class detection. The proposed model based on stacked-Dense-CNN cascade architecture is robust and resilient to the class imbalance and has better overall detection accuracy. The first stage of the stacked-Dense-CNN cascade consists of an optimal feature learning stage, comprising a LASSO (least absolute shrinkage and selection) and majority voting step, for extraction of significant and homogenized features. The second stage use of a novel stacked-Dense-CNN cascade model and a novel model development protocol involving an unique train-test dataset partitioning strategy. Also, by using a specific training routine per epoch, similar to the simulated annealing approach, it was possible to achieve enhanced detection performance, particularly for detection of minority class, and robustness to class imbalance. The experimental evaluation of the novel stacked-Dense-CNN cascade model on a super dataset obtained by fusing multiple data subsets of publicly available NHANES data, resulted in an accuracy of 81.8% accuracy for negative CVD cases (majority class), and 85% for the positive CVD cases (minority class), an improved performance as compared to previously proposed research approaches for imbalanced clinical data settings. </p
Universal Object Detection Under Unconstrained Environments
No full textThis paper presents a universal object detection framework for unconstrained environment settings where machines can only learn from massive unlabeled multimodal data and a few labeled data. This research aims to tackle key challenges in computer vision and expects to produce next-generation object detection techniques that can effectively detect objects of diversified categories in complex application settings. The proposed universal object detection framework is based on a novel formulation to solve anomaly detection problem leveraging multimodal heterogeneous data sources and denoising diffusion models and application to a wide set of complex application settings.</p
Toward a Generic Multi-modal Medical Data Representation Model
No full textThis paper presents a generic multi-modal medical data representation model, based on utilizing the knowledge from the abundance of medical data in the publicly available medical imaging databases. Using novel deep learning techniques, this paper proposes an AI model that is technically capable of capturing characteristics of complex health conditions. The findings from this work based on extensive experimental work allows the development of robust and automatic detection of gliomas, an aggressive form of brain tumors. This model can provide significant benefits to the wide medical AI community and stimulate development and benefit universal health care in the long term.</p
Semantic Segmentation and Pathology Localization in Lung Ultrasound Images Using Transfer Learning
No full textSignificant progress has been made to leverage machine learning towards decision support diagnosis from ultrasound in medical imaging. Due to the continuous challenges associated with evolving characteristics of COVID-19, the availability of fast, safe, and highly sensitive diagnostic tools is imperative. Where ultrasound imaging has been proven superior to X-rays and CT scans, the main limitation to its use is operator dependency and experience. In this study, we propose a novel transfer learning and semantic segmentation framework for automatic detection and localization of multiple lung pathologies in Lung Ultrasound images. The proposed framework allows better interpretation and explanation of the model decisions, with clear visualization and localization of the different pathologies. The experimental evaluation of the proposed framework was done on an open-source Lung ultrasound imaging dataset which is labeled and annotated by team of radiologists. The proposed approach was validated using other benchmarks models for comparison in terms of DICE coefficients at 0.98 and IOU Score at 0.97 which outperforms the other benchmark models quite significantly. This provides interpretation of the reasoning behind the decision made by the model leading to higher rates of acceptance with clinicians
Intelligent human activity recognition scheme for health applications
No full textAutomatic activity recognition systems aim to capture the state of the user and its environment by exploiting heterogeneous sensors, and permit continuous monitoring of numerous physiological signals, where these sensors are attached to the subject\u27s body. This can be immensely useful in healthcare applications, for automatic and intelligent daily activity monitoring for elderly people. In this paper, we present a novel data analytic scheme for intelligent Human Activity Recognition (AR) using wireless body sensors and smartphone inertial sensors which use information theory-based feature ranking algorithms and classifiers based on random forests, ensemble learning and lazy learning. Further, we propose a novel multimodal scheme based on combining multimodal three dimensional (x, y, z) accelerometer and gyro data from smart phone inertial sensors. Extensive experiments using different publicly available database of human activity show that the proposed approach can assist in the development of intelligent and automatic real time human activity monitoring technology for eHealth application scenarios for elderly, disabled and people with special needs
Multiview gait biometrics for human identity recognition
No full textWe propose a novel multiview fusion scheme for recognizing human identity based on gait biometric data. The gait biometric data is acquired from video surveillance datasets from multiple cameras. Experiments on publicly available CASIA dataset show the potential of proposed scheme based on fusion towards development and implementation of automatic identity recognition systems
Message from DSS 2020 General Chairs
No full textMessage from the DSS 2020 General Chairs HPCC-SmartCity-DSS 2020 Welcome to the 6th IEEE International Conference on Data Science and Systems (DSS 2020). Given the COVID-19 pandemic and associated travel restrictions, as the safety of people is of the highest priority, the conference will be held virtually on December 14-16, 2020. On behalf of the Organizing Committee of DSS 2020, we would like to express our sincere and warm welcome to all of participants! The IEEE DSS 2020 Conference, the 6th event in the series, a prime international forum for researchers, industry practitioners and domain experts to exchange the latest advances in Data Science and Data Systems as well as their synergy. DSS 2020 is sponsored by IEEE, IEEE Computer Society, IEEE Technical Committee on Scalable Computing (TCSC), and IEEE CPSS. DSS 2020 consisted of the main conference with 12 regular and 6 short paper presentations out of 69 submissions from more than 10 countries or regions. For the successful initialization and organization of this international conference with this size and diversity, we counted on the great support of many people and organizations. First of all, we would like to sincerely thank Laurence T. Yang (St. Francis Xavier University, Canada) and Jinjun Chen (Swinburne University of Technology, Australia), the Steering Chairs of DSS, for giving us the opportunity to organize the conference and for their support and guidance. We would like to express our special thanks to the Program Chairs Gunasekaran Manogaran (University of California, Davis, USA), Francesco Piccialli (University of Naples FEDERICO II, Italy), and Yaliang Zhao (St. Francis Xavier University, Canada & Henan University, China) for their excellent work and tremendous efforts in organizing an excellent program committee, conducting a rigorous double-blind review, selecting high quality papers from a large number of submissions, and preparing an excellent conference. We are grateful to the Workshop Chairs Jinke Wang (Henan University, China), Leo Y. Zhang (Deakin University, Australia), Jie Lei (Nanchang University, China), and Caihong Yuan (Henan University, China), as well as other chairs and members for their great supports. We thank all the reviewers for their hard work in reviewing the manuscript, providing constructive feedback to the authors and making the paper well selected. Most importantly, we are grateful to all the authors for the high quality of the papers submitted to the main DSS 2020 conference and its workshops. Last but not least, we would like to thank the DSS 2020 web and the Virtual Conference organizing team for the excellent arrangements of the conference. Thank you to everyone who attended DSS 2020, we hope the conference will be exciting and interesting for your research and professional activities, and that IEEE DSS will be one of the best conferences in the field! Michael Sheng, Macquarie University, Australia Girija Chetty, University of Canberra, Australia Xiaowen Chu, Hong Kong Baptist University, Hong Kong DSS 2020 General Chair
- …
