2,997 Hits in 1.4 sec

A ConvNet for the 2020s [article]

Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie
2022 arXiv   pre-print
In this work, we reexamine the design spaces and test the limits of what a pure ConvNet can achieve.  ...  The outcome of this exploration is a family of pure ConvNet models dubbed ConvNeXt.  ...  We thank Kaiming He, Eric Mintun, Xingyi Zhou, Ross Girshick, and Yann LeCun for valuable discussions and feedback.  ... 
arXiv:2201.03545v2 fatcat:eelgnxxjorgizp7fxkpqo74osa

RealHePoNet: a robust single-stage ConvNet for head pose estimation in the wild [article]

Rafael Berral-Soler, Francisco J. Madrid-Cuevas, Rafael Muñoz-Salinas, Manuel J. Marín-Jiménez
2020 arXiv   pre-print
As a result of this work, we have obtained a trained ConvNet model, coined RealHePoNet, that given a low-resolution grayscale input image, and without the need of using facial landmarks, is able to estimate  ...  In this work, we address this problem, defined here as the estimation of both vertical (tilt/pitch) and horizontal (pan/yaw) angles, through the use of a single Convolutional Neural Network (ConvNet) model  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan X Pascal GPU used for this research.  ... 
arXiv:2011.01890v1 fatcat:eym2rlg4ajhgvpx5qzkdlqmhtq

Biotic Yield Losses in the Southern Amazon, Brazil: Making Use of Smartphone-Assisted Plant Disease Diagnosis Data

Anna C. Hampf, Claas Nendel, Simone Strey, Robert Strey
2021 Frontiers in Plant Science  
Between 2016 and 2020, Plantix users collected approximately 78,000 georeferenced P&A images in the Southern Amazon.  ...  The study results indicate a high performance of the trained ConvNets in classifying 420 different crop-disease combinations.  ...  We were thankful for the helpful comments and suggestions made by the reviewers.  ... 
doi:10.3389/fpls.2021.621168 pmid:33936124 pmcid:PMC8083370 fatcat:bjlhq4cvpfazxfi6ro6gl5u7fy

Medical Image Segmentation via Unsupervised Convolutional Neural Network [article]

Junyu Chen, Eric C. Frey
2020 arXiv   pre-print
For the majority of the learning-based segmentation methods, a large quantity of high-quality training data is required.  ...  Specifically, in the unsupervised setting, we parameterize the Active contour without edges (ACWE) framework via a convolutional neural network (ConvNet), and optimize the parameters of the ConvNet using  ...  Acknowledgments This work was supported by a grant from the National Cancer Institute, U01-CA140204.  ... 
arXiv:2001.10155v4 fatcat:6dfunjgrnvhi3fgcrrqw7lgjm4

Differentiating COVID-19 from other types of pneumonia with convolutional neural networks [article]

Ilker Ozsahin, Confidence Onyebuchi, Boran Sekeroglu
2020 medRxiv   pre-print
INTRODUCTION A widely-used method for diagnosing COVID-19 is the nucleic acid test based on real-time reverse transcriptase-polymerase chain reaction (RT-PCR).  ...  CONCLUSIONS The ConvNet was able to distinguish the COVID-19 images among non-COVID-19 images, namely bacterial and viral pneumonia as well as normal X-ray images.  ...  The potential for the virus to spread widely resulted in a global pandemic in 2020, which caused the World Health Organization (WHO) to declare a public health emergency of international concern (PHEIC  ... 
doi:10.1101/2020.05.26.20113761 fatcat:otrfn57qqjf2hdpf6hzjf3lzem

Multi-Channel ConvNet Approach to Predict the Risk of In-Hospital Mortality for ICU Patients [article]

Mahmoud Elbattah
2020 Figshare  
The key idea is to disaggregate multi-variate TS into separate channels, where a ConvNet is used to extract features from each univariate TS individually.  ...  Our experimental results show a promising accuracy of classification that is competitive to the state-of-the-art.  ...  datasets and multiple variables. • For example, ConvNets were applied as a feature extractor for multivariate TS classification (Zheng et al. 2016 ). • RNN architectures were also explored (e.g.  ... 
doi:10.6084/m9.figshare.12712181 fatcat:62ww2vmrt5hezmlghqqdsbuehq

EdgeFormer: Improving Light-weight ConvNets by Learning from Vision Transformers [article]

Haokui Zhang, Wenze Hu, Xiaoyu Wang
2022 arXiv   pre-print
We propose EdgeFormer, a pure ConvNet based backbone model that further strengthens these advantages by fusing the merits of vision transformers into ConvNets.  ...  However, in the area of small models for mobile or resource constrained devices, ConvNet still has its own advantages in both performance and model complexity.  ...  et al. (2020) .  ... 
arXiv:2203.03952v2 fatcat:zhrlsurr6vhjfdpehqtajnhrey

Cellular-Neural-Network Focal-Plane Processor as Pre-Processor for ConvNet Inference

Lionel C. Gontard, Ricardo Carmona-Galan, Angel Rodriguez-Vazquez
2020 2020 IEEE International Symposium on Circuits and Systems (ISCAS)  
In this paper we carry out some experiments on the implementation of ConvNets with CNN hardware in the form of a focal-plane image processor.  ...  It is shown that ultra-fast inference can be implemented, using as an example a LeNetbased ConvNet architecture.  ...  ACKNOWLEDGMENTS The authors want to acknowledge funding from MCIU/AEI/ERDF-EU through projects PGC2018-101538-A-I00 and RTI2018-097088-B-C31.  ... 
doi:10.1109/iscas45731.2020.9181102 fatcat:iorpuc5cazhgtgpjwjt5okekxm

VersaTile Convolutional Neural Network Mapping on FPGAs

A. Munio-Gracia, J. Fernandez-Berni, R. Carmona-Galan, A. Rodriguez-Vazquez
2020 2020 IEEE International Symposium on Circuits and Systems (ISCAS)  
Convolutional Neural Networks (ConvNets) are directed acyclic graphs with node transitions determined by a set of configuration parameters.  ...  In this paper, we describe a dynamically configurable hardware architecture that enables data allocation strategy adjustment according to ConvNets layer characteristics.  ...  Thus full system configuration for a 26-layer ConvNet as SqueezeNet v1.1 can be transferred in 780 ns at a 200 MHz clock frequency. IV.  ... 
doi:10.1109/iscas45731.2020.9181037 fatcat:44a37wqqevbg5ed4lggytazcui

Filtering Internal Tides From Wide-Swath Altimeter Data Using Convolutional Neural Networks [article]

Redouane Lguensat, Ronan Fablet, Julien Le Sommer, Sammy Metref, Emmanuel Cosme, Kaouther Ouenniche, Lucas Drumetz, Jonathan Gula
2020 arXiv   pre-print
The upcoming Surface Water Ocean Topography (SWOT) satellite altimetry mission is expected to yield two-dimensional high-resolution measurements of Sea Surface Height (SSH), thus allowing for a better  ...  In this paper, we cast this problem into a supervised learning framework and propose the use of convolutional neural networks (ConvNets) to estimate fields free of internal tide signals.  ...  SWOT 2020 IEEE. Personal use of this material is permitted.  ... 
arXiv:2005.01090v1 fatcat:fgrgpg6rz5fd5c5cqi3u53ezwa

Sensor Classification Using Convolutional Neural Network by Encoding Multivariate Time Series as Two-Dimensional Colored Images

Chao-Lung Yang, Zhi-Xuan Chen, Chen-Yi Yang
2019 Sensors  
The framework encodes multivariate time series data into two-dimensional colored images, and concatenate the images into one bigger image for classification through a Convolutional Neural Network (ConvNet  ...  Surprisingly, the simple structure of ConvNet is sufficient enough for classification as it performed equally well with the complex structure of VGGNet.  ...  The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.  ... 
doi:10.3390/s20010168 pmid:31892141 pmcid:PMC6982717 fatcat:ib33nkidnresngaedqhfa2pcfy

Learning a Lie Algebra from Unlabeled Data Pairs [article]

Christopher Ick, Vincent Lostanlen
2020 arXiv   pre-print
For example, the disentanglement of pitch, intensity dynamics, and playing technique remains a challenging task in music information retrieval.  ...  Deep convolutional networks (convnets) show a remarkable ability to learn disentangled representations.  ...  arXiv:2009.09321v2 [cs.LG] 22 Sep 2020  ... 
arXiv:2009.09321v3 fatcat:hx6s4w4a25bmvk5hgndgnmth5a

Handwritten Phoenician Character Recognition and its Use to Improve Recognition of Handwritten Alphabets with Lack of Annotated Data

Lamyaa Sadouk
2020 International Journal of Advanced Trends in Computer Science and Engineering  
To this matter, a database for Phoenician handwritten characters (PHCDB) is introduced for the first time in this paper.  ...  As such, the availability of a reference database for Phoenician handwritten characters is crucial to carry out these tasks.  ...  Figure 6 : 6 The 5 by 5 filter weights of the 1st convolutional layer for the following approaches: (a) Phoenician ConvNet, (b) Tifinagh ConvNet, (c) Full TL using Phoenician ConvNet.  ... 
doi:10.30534/ijatcse/2020/26912020 fatcat:ysghomrbn5hppnau2vou4n4bgy

An End to End Framework with Adaptive Spatio-Temporal Attention Module for Human Action Recognition

Shaocan Liu, Xin Ma, Hanbo Wu, Yibin Li
2020 IEEE Access  
For more information, see VOLUME 8, 2020  ...  Spatial Attention Module is established by fusing the value feature and the gradient feature of the feature map, making the representation of ConvNet for action recognition focus on the informative motion  ...  VOLUME 8, 2020 First convolutional neural networks for action recognition, then attention mechanisms. A.  ... 
doi:10.1109/access.2020.2979549 fatcat:4ws544xeifbutkl5yed5anyntq

Bag of Tricks for Retail Product Image Classification [article]

Muktabh Mayank Srivastava
2020 arXiv   pre-print
These tricks enable us to increase the accuracy of fine tuned convnets for retail product image classification by a large margin.  ...  As the most prominent trick, we introduce a new neural network layer called Local-Concepts-Accumulation (LCA) layer which gives consistent gains across multiple datasets.  ...  In this work, we show methods to improve finetuning [6] of convnets for image classification of retail objects. arXiv:2001.03992v1 [cs.CV] 12 Jan 2020 Convnets have been shown to work very well for image  ... 
arXiv:2001.03992v1 fatcat:es6ew55tknfqhdg7z5dqlklysq
« Previous Showing results 1 — 15 out of 2,997 results