Filters








15,449 Hits in 10.9 sec

Error Corrective Boosting for Learning Fully Convolutional Networks with Limited Data [article]

Abhijit Guha Roy, Sailesh Conjeti, Debdoot Sheet, Amin Katouzian, Nassir Navab, Christian Wachinger
2017 arXiv   pre-print
For the subsequent fine-tuning of the network with manually labeled data, we introduce error corrective boosting (ECB), which emphasizes parameter updates on classes with lower accuracy.  ...  Training deep fully convolutional neural networks (F-CNNs) for semantic image segmentation requires access to abundant labeled data.  ...  The authors would also like to thank Magdalini Paschali for proof reading and feedback.  ... 
arXiv:1705.00938v2 fatcat:bsjgb65bmjgj5jvyk6lr6dmol4

Deep Boosted Regression for MR to CT Synthesis [chapter]

Kerstin Kläser, Pawel Markiewicz, Marta Ranzini, Wenqi Li, Marc Modat, Brian F. Hutton, David Atkinson, Kris Thielemans, M. Jorge Cardoso, Sébastien Ourselin
2018 Lecture Notes in Computer Science  
We propose a novel deep fully convolutional neural network that generates synthetic CTs in a recursive manner by gradually reducing the residuals of the previous network, increasing the overall accuracy  ...  Attenuation correction is an essential requirement of positron emission tomography (PET) image reconstruction to allow for accurate quantification.  ...  Acknowledgements This work was supported by an IMPACT studentship funded jointly by Siemens and the EPSRC UCL Centre for Doctoral Training in Medical Imaging (EP/L016478/1).  ... 
doi:10.1007/978-3-030-00536-8_7 fatcat:j3ylahzfirh5lnxwqspqig3u3e

Deep Incremental Boosting [article]

Alan Mosca, George D Magoulas
2017 arXiv   pre-print
This paper introduces Deep Incremental Boosting, a new technique derived from AdaBoost, specifically adapted to work with Deep Learning methods, that reduces the required training time and improves generalisation  ...  We show a set of experiments that outlines some preliminary results on some common Deep Learning datasets and discuss the potential improvements Deep Incremental Boosting brings to traditional Ensemble  ...  convolutions, with 25% dropout • 256 3 × 3 convolutions, with 25% dropout • 2 × 2 max-pooling • A fully connected layer of 1024 nodes, with 50% dropout • a Softmax layer with 10 outputs (one for  ... 
arXiv:1708.03704v1 fatcat:pyr5jzchcfedhpliiqjp62ax4m

n Artificial Intelligence Approach Based on Hybrid CNN-XGB Model to Achieve High Prediction Accuracy through Feature Extraction, Classification and Regression for Enhancing Drug Discovery in Biomedicine

Mukesh Madanan, Biju T. Sayed, Nurul Akhmal Mohd Zulkefli, Nitha C. Velayudhan
2021 International Journal of Biology and Biomedical Engineering  
as the input data of the XGBoost for drug response prediction.  ...  Herein, the paper proposes a deep neural network structure as the Convolutional Neural Network (CNN) to detain the gene expression features of the cell line and then use the resulting abstract features  ...  Then, by training the Convolutional Neural Network model along with output layer of Fully Connected (FC), the training error is fixed.  ... 
doi:10.46300/91011.2021.15.22 fatcat:iu6ujquoarba3aqhph32s5pina

Boosting Information Extraction Systems with Character-level Neural Networks and Free Noisy Supervision

Philipp Meerkamp, Zhengyi Zhou
2017 Proceedings of the 2nd Workshop on Structured Prediction for Natural Language Processing  
Our architecture combines the ability of constraint-based or hybrid extraction systems to easily incorporate domain knowledge with the ability of deep neural networks to leverage large amounts of data  ...  to learn complex features.  ...  Acknowledgments We would like to thank my managers Alex Bozic, Tim Phelan, and Joshwini Pereira for supporting this project, as well as David Rosenberg from the CTO's office for providing access to GPU  ... 
doi:10.18653/v1/w17-4307 dblp:conf/emnlp/MeerkampZ17 fatcat:zehu3g7zrbdwlckwrj5p76d7ru

Deep Convolutional Neural Network Ensembles Using ECOC

Sara Atito Ali Ahmed, Cemre Zor, Muhammad Awais, Berrin Yanikoglu, Josef Kittler
2021 IEEE Access  
In this paper, we analyse an error correcting output coding (ECOC) framework for constructing ensembles of deep networks and propose different design strategies to address the accuracy-complexity trade-off  ...  We carry out an extensive comparative study between the introduced ECOC designs and the state-of-the-art ensemble techniques such as ensemble averaging and gradient boosting decision trees.  ...  error correction.  ... 
doi:10.1109/access.2021.3088717 fatcat:yk6uitl6o5f77iutveserzmle4

Reaching Human-level Performance in Automatic Grammatical Error Correction: An Empirical Study [article]

Tao Ge, Furu Wei, Ming Zhou
2018 arXiv   pre-print
Fluency boosting learning generates diverse error-corrected sentence pairs during training, enabling the error correction model to learn how to improve a sentence's fluency from more instances, while fluency  ...  Combining fluency boost learning and inference with convolutional seq2seq models, our approach achieves the state-of-the-art performance: 75.72 (F_0.5) on CoNLL-2014 10 annotation dataset and 62.42 (GLEU  ...  Fluency boost learning fully exploits both errorcorrected data and native data by generating diverse error-corrected sentence pairs during training, which benefits model learning and improves the performance  ... 
arXiv:1807.01270v5 fatcat:gxofu3amg5dxtnif6sli7oqlue

Abstract: Fast MRI Whole Brain Segmentation with Fully Convolutional Neural Networks [chapter]

Abhijit Guha Roy, Sailesh Conjeti, Nassir Navab, Christian Wachinger
2018 Bildverarbeitung für die Medizin 2018  
Error corrective boosting for learning fully convolutional networks with limited data. Springer 2017. 2. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain.  ...  We introduced a fully convolution neural network (F-CNN) that segments a brain scan in several seconds [1] .  ...  We introduced a fully convolution neural network (F-CNN) that segments a brain scan in several seconds [1] .  ... 
doi:10.1007/978-3-662-56537-7_26 dblp:conf/bildmed/RoyCNW18 fatcat:2hl7gdfyazbsdknyzqj3oceyom

Fusion of Heterogeneous Data in Convolutional Networks for Urban Semantic Labeling (Invited Paper) [article]

Nicolas Audebert , Sébastien Lefèvre
2017 arXiv   pre-print
In this work, we present a novel module to perform fusion of heterogeneous data using fully convolutional networks for semantic labeling.  ...  We introduce residual correction as a way to learn how to fuse predictions coming out of a dual stream architecture.  ...  ACKNOWLEDGMENT The Vaihingen dataset was provided by the German Society for Photogrammetry, Remote Sensing and Geoinformation (DGPF) [15] : http://www.ifp.uni-stuttgart.de/dgpf/DKEP-Allg. html.  ... 
arXiv:1701.05818v1 fatcat:o7o3khz3mbgpplxcdvsbfrt7pi

Deep Convolutional Neural Network Ensembles using ECOC [article]

Sara Atito Ali Ahmed, Cemre Zor, Berrin Yanikoglu, Muhammad Awais, Josef Kittler
2021 arXiv   pre-print
In this paper, we analyse error correcting output coding (ECOC) framework to be used as an ensemble technique for deep networks and propose different design strategies to address the accuracy-complexity  ...  We carry out an extensive comparative study between the introduced ECOC designs and the state-of-the-art ensemble techniques such as ensemble averaging and gradient boosting decision trees.  ...  error correction.  ... 
arXiv:2009.02961v2 fatcat:6avvpsnpgrbtblnboxbuhkjy5u

A Combined Deep Learning-Gradient Boosting Machine Framework for Fluid Intelligence Prediction [chapter]

Yeeleng S. Vang, Yingxin Cao, Xiaohui Xie
2019 Lecture Notes in Computer Science  
We train a convolutional neural network to compress the high dimensional MRI data and learn meaningful image features by predicting the 123 continuous-valued derived data provided with each MRI.  ...  In this work, we propose a deep learning combined with gradient boosting machine framework to solve this task.  ...  Specifically we design a fully convolutional network to perform data compression/feature extraction.  ... 
doi:10.1007/978-3-030-31901-4_1 fatcat:fuknspmndzay3oqchyghjaarcq

Arabic Handwritten Character Recognition based on Convolution Neural Networks and Support Vector Machine

Mahmoud Shams, Amira. A., Wael. Z.
2020 International Journal of Advanced Computer Science and Applications  
In this paper, we present an algorithm for recognizing Arabic letters and characters based on using deep convolution neural networks (DCNN) and support vector machine (SVM).  ...  Moreover, we determine the error classification rate (ECR).  ...  The proposed system achieves 95.07% correct classification accuracy with a minimum error classification rate of 4.93% compared with recent approaches.  ... 
doi:10.14569/ijacsa.2020.0110819 fatcat:qsct3ixx3zh6jjtk3bi3xghlka

Learning to Count with CNN Boosting [chapter]

Elad Walach, Lior Wolf
2016 Lecture Notes in Computer Science  
We follow modern learning approaches in which a density map is estimated directly from the input image.  ...  We employ CNNs and incorporate two significant improvements to the state of the art methods: layered boosting and selective sampling.  ...  Acknowledgments This research is supported by the Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI). Learning to Count with CNN Boosting  ... 
doi:10.1007/978-3-319-46475-6_41 fatcat:d74atjc4yjgbphwtf42vxwj4j4

Semantic Segmentation of Earth Observation Data Using Multimodal and Multi-scale Deep Networks [article]

Nicolas Audebert , Sébastien Lefèvre
2016 arXiv   pre-print
This work investigates the use of deep fully convolutional neural networks (DFCNN) for pixel-wise scene labeling of Earth Observation images.  ...  predictions at multiple scales; 3) we perform data fusion from heterogeneous sensors (optical and laser) using residual correction.  ...  The Vaihingen data set was provided by the German Society for Photogrammetry, Remote Sensing and Geoinformation (DGPF) [39] : http://www.ifp.uni-stuttgart.de/dgpf/DKEP-Allg.html.  ... 
arXiv:1609.06846v1 fatcat:7tu6as23pbd7xgsnjzokor3bp4

Semantic Segmentation of Earth Observation Data Using Multimodal and Multi-scale Deep Networks [chapter]

Nicolas Audebert, Bertrand Le Saux, Sébastien Lefèvre
2017 Lecture Notes in Computer Science  
This work investigates the use of deep fully convolutional neural networks (DFCNN) for pixel-wise scene labeling of Earth Observation images.  ...  predictions at multiple scales; 3) we perform data fusion from heterogeneous sensors (optical and laser) using residual correction.  ...  The Vaihingen data set was provided by the German Society for Photogrammetry, Remote Sensing and Geoinformation (DGPF) [39] : http://www.ifp.uni-stuttgart.de/dgpf/DKEP-Allg.html.  ... 
doi:10.1007/978-3-319-54181-5_12 fatcat:k4kpmewrrfetdp36rq3xrluma4
« Previous Showing results 1 — 15 out of 15,449 results