Filters








11 Hits in 4.8 sec

CardioXNet: A Novel Lightweight Deep Learning Framework for Cardiovascular Disease Classification Using Heart Sound Recordings

Samiul Based Shuvo, Shams Nafisa Ali, Soham Irtiza Swapnil, Mabrook S. Al-Rakhami, Abdu Gumaei
2021 IEEE Access  
INDEX TERMS Phonocardiogram analysis, unsegmented heart sound, cardiovascular disease, lightweight CRNN architecture, deep learning, SqueezeNet.  ...  In this article, we propose CardioXNet, a novel lightweight end-to-end CRNN architecture for automatic detection of five classes of cardiac auscultation namely normal, aortic stenosis, mitral stenosis,  ...  PERFORMANCE OF THE PROPOSED FRAMEWORK In this work, the proposed lightweight model has been evaluated on the Github PCG datast, PhysioNet/CinC challenge dataset and on both of these dataset combined.  ... 
doi:10.1109/access.2021.3063129 fatcat:io5rva7lnnay5hkuft6sllqkz4

Federated Learning for Intrusion Detection System: Concepts, Challenges and Future Directions [article]

Shaashwat Agrawal, Sagnik Sarkar, Ons Aouedi, Gokul Yenduri, Kandaraj Piamrat, Sweta Bhattacharya, Praveen Kumar Reddy Maddikunta, Thippa Reddy Gadekallu
2021 arXiv   pre-print
On the contrary, federated learning (FL) fits in appropriately as a privacy-preserving decentralized learning technique that does not transfer data but trains models locally and transfers the parameters  ...  Machine Learning and Deep Learning with Intrusion Detection Systems have gained great momentum due to their achievement of high classification accuracy.  ...  Thirdly, the use of deep learning models which have enabled self-learning ensuring optimized accuracy. But there exists several challenges associated with data privacy and security.  ... 
arXiv:2106.09527v1 fatcat:vsy4l2ew4nbh5j3gzdlao4ngxe

Deep learning approach to control of prosthetic hands with electromyography signals [article]

Mohsen Jafarzadeh, Daniel Curtiss Hussey, Yonas Tadesse
2019 arXiv   pre-print
In this paper, we propose a deep learning approach to control prosthetic hands with raw EMG signals. We use a novel deep convolutional neural network to eschew the feature-engineering step.  ...  The proposed approach is implemented in Python with TensorFlow deep learning library, and it runs in real-time in general-purpose graphics processing units of NVIDIA Jetson TX2 developer kit.  ...  John Hanson, for comments and guidance that greatly improved the projects. We would also like to show our gratitude to Dr.  ... 
arXiv:1909.09910v1 fatcat:j5g6kir3gbelnh7lnww3xlyaqm

Applications of Artificial Intelligence in Battling Against Covid-19: A Literature Review

Mohammad-H. Tayarani-N.
2020 Chaos, Solitons & Fractals  
The rapid growth of the pandemic has wreaked havoc and prompted the need for immediate reactions to curb the effects.  ...  We also tried to conclude the paper with ideas on how the problems can be tackled in a better way and provide some suggestions for future works.  ...  A lightweight deep learning algorithm is proposed in [161] . The algorithm is used to perform seg-mentation on covid-19 CT images.  ... 
doi:10.1016/j.chaos.2020.110338 pmid:33041533 pmcid:PMC7532790 fatcat:gl3i37hag5gflajsa7fh6khvva

SHREC 2020 Track: 6D Object Pose Estimation

Honglin Yuan, Remco C. Veltkamp, Georgios Albanis, Nikolaos Zioulis, Dimitrios Zarpalas, Petros Daras
2020 Eurographics Workshop on 3D Object Retrieval, EG 3DOR  
At the same time, existing 3D datasets that are used for data-driven methods to estimate 6D poses have limited view angles and low resolution.  ...  Data-driven methods are the current trend in 6D object pose estimation and our evaluation results show that approaches which fully exploit the color and geometric features are more robust for 6D pose estimation  ...  Instead of relying on improving handcrafted features, they learn more robust features and semantic cues by applying deep learning models.  ... 
doi:10.2312/3dor.20201164 fatcat:tdwgzolgtrdrxpuhge6rnndlb4

Task-Adaptive Neural Network Search with Meta-Contrastive Learning [article]

Wonyong Jeong, Hayeon Lee, Gun Park, Eunyoung Hyung, Jinheon Baek, Sung Ju Hwang
2021 arXiv   pre-print
Given a model-zoo that consists of network pretrained on diverse datasets, we use a novel amortized meta-learning framework to learn a cross-modal latent space with contrastive loss, to maximize the similarity  ...  parameters), from a model zoo.  ...  still meaningfully low, which implies that our performance model successfully works even toward the entire model-zoo.  ... 
arXiv:2103.01495v2 fatcat:rc2fk4eb35bj5ed4ozgsnvh73e

Spatiotemporal enabled Content-based Image Retrieval

Mariana Belgiu, Martin Sudmanns, Tiede Dirk, Andrea Baraldi, Stefan Lang
2016 International Conference on GIScience Short Paper Proceedings  
For efficient deployment of sensors in a WSN the coverage estimation is a critical issue. Probabilistic methods are among the most accurate models proposed for sensor coverage estimation.  ...  This is because raster representations are constrained by their spatial resolution, and their regular shapes result in redundant data for unoccupied areas.  ...  We are confident that our experiments will efficiently learn the optimal parameters, and thus improve the estimation accuracy of the interpolation model, helping us to definitively establish more accurate  ... 
doi:10.21433/b311729295dw fatcat:fulw4pw3kfh5nmfzcsy3pkisvm

From Manual to Automated Design of Biomedical Semantic Segmentation Methods

Fabian Isensee
2021
First and foremost I would like to express my gratitude towards my supervisor, Klaus Maier-Hein for his continued guidance and encouragement.  ...  project should move next and what research directions will be relevant in the future are unmatched -it is thanks to his foresight that we took it upon us to develop robust and generalizable segmentation models  ...  The performance of a machine learning model, in particular those that use non-deep 20 learning methods (for example the methods described in Section 2. 1.3 ) strongly depends on the availability of a  ... 
doi:10.11588/heidok.00029345 fatcat:zuudfkljrfaz5nng6jggbzqygm

Design Space Exploration of Data-centric Architectures [article]

Smriti Prathapan, Maryland Shared Open Access Repository, Milton ; Halem
2021
We define a generic NDP architecture which is well-suited for memory-bound computations and implement the software kernels for NDP-based algorithmic mapping.We show for a modest sized NDP system, that  ...  big data? is leading to changes in the compute paradigm, in particular to the notion of moving computation to data, known as Near Data Processing (NDP).  ...  It is an instance-based or a lazy learning method which performs the learning process at the time when new sample is to be classified as opposed to other learning models where the training data is pre-classified  ... 
doi:10.13016/m2ctv2-wote fatcat:smh76fzd6rdwjltajd56qbnoea

Proceedings of the Seminars Future Internet (FI) and Innovative Internet Technologies and Mobile Communication (IITM), Summer Semester 2018 [article]

Georg Carle, Daniel Raumer, Stephan Günther, Benedikt Jaeger, Chair Of Network Architectures
2018
Deep Learning gained in popularity, when deep convolutional networks performed extraordinarily well on the ImageNet challenge in 2012 [10] .  ...  Nevertheless it is very likely that these deep learning frameworks will adapt with time.  ...  Keywords Distributed Systems, Data Management, Peer-To-Peer, Content Delivery Network, Scalability, Performance, Consistency, Redundancy, Overhead  ... 
doi:10.2313/net-2018-11-1 fatcat:bnh7d4o7pna4njsnu52zvjcsou

Institute of Geographic Science and Natural Resources Research

Jinfeng Wang, Full Phd, Professor
Chinese Academy of Sciences   unpublished
The theoretical models for teaching and learning provide the geography teacher with the so-called "backbone" for every lesson.  ...  Velocity modeling for time-depth conversion involves building velocity model using all available velocity data.  ... 
fatcat:sahplddiijadrflo3csqtw5mfu