8,478 Hits in 5.0 sec

Deep Neural Network for DrawiNg Networks, (DNN)^2 [article]

Loann Giovannangeli, Frederic Lalanne, David Auber, Romain Giot, Romain Bourqui
2021 arXiv   pre-print
In this paper, we present a novel graph drawing framework called (DNN)^2: Deep Neural Network for DrawiNg Networks. Our method uses Graph Convolution Networks to learn a model.  ...  The results show that (DNN)^2 performs well and are encouraging as the Deep Learning approach to Graph Drawing is novel and many leads for future works are identified.  ...  This paper presents Deep Neural Networks for DrawiNg Networks, (DNN) 2 , a graph layout framework relying on unsupervised Deep Learning.  ... 
arXiv:2108.03632v2 fatcat:4k2rmxh7lrh5rfaiughnx5phnq

Faster Convergence & Generalization in DNNs [article]

Gaurav Singh, John Shawe-Taylor
2018 arXiv   pre-print
Deep neural networks have gained tremendous popularity in last few years. They have been applied for the task of classification in almost every domain.  ...  Despite the success, deep networks can be incredibly slow to train for even moderate sized models on sufficiently large datasets.  ...  Related Works Deep neural networks started to gain popularity in late 90's [6, 10, 9] .  ... 
arXiv:1807.11414v3 fatcat:3ryn545btvbjlhkmwgf5cqa3yu

Weight Initialization of Deep Neural Networks(DNNs) using Data Statistics [article]

Saiprasad Koturwar, Shabbir Merchant
2018 arXiv   pre-print
Deep neural networks (DNNs) form the backbone of almost every state-of-the-art technique in the fields such as computer vision, speech processing, and text analysis.  ...  We think the reason for this is not making use of data statistics for initializing the network weights.  ...  Introduction Deep neural networks(DNN) are being used very heavily in the fields such as computer vision [1] , speech processing [2] and text processing [3] .  ... 
arXiv:1710.10570v2 fatcat:neiimtbbwbgcngx2zbydnad4xa

Dynamic Network Surgery for Efficient DNNs [article]

Yiwen Guo, Anbang Yao, Yurong Chen
2016 arXiv   pre-print
Deep learning has become a ubiquitous technology to improve machine intelligence.  ...  In this paper, we propose a novel network compression method called dynamic network surgery, which can remarkably reduce the network complexity by making on-the-fly connection pruning.  ...  The recently proposed BinaryConnect [2] and Binarized Neural Networks [3] are able to compress DNNs by a factor of 32×, while a noticeable accuracy loss is sort of inevitable.  ... 
arXiv:1608.04493v2 fatcat:3truyx2r4vdylif6sa5rpjrwwy

Visualizing Global Explanations of Point Cloud DNNs [article]

Hanxiao Tan
2022 arXiv   pre-print
So far, however, there has been little discussion about the explainability of deep neural networks for point clouds.  ...  Therefore, point cloud neural networks have become a popular research direction in recent years.  ...  Nevertheless, point clouds are structurally different from traditional image deep neural networks (DNN)s so that the aforementioned AM methods are not directly applicable to point cloud networks.  ... 
arXiv:2203.09505v2 fatcat:o33sf64nwbfknhgw3r2b4cgzey

Improving DNN Robustness to Adversarial Attacks using Jacobian Regularization [article]

Daniel Jakubovitz, Raja Giryes
2019 arXiv   pre-print
Deep neural networks have lately shown tremendous performance in various applications including vision and speech processing tasks.  ...  For this reason, providing robustness to adversarial attacks is an important challenge in networks training, which has led to extensive research.  ...  Introduction Deep neural networks (DNNs) are a widespread machine learning technique, which has shown state-of-the-art performance in many domains such as natural language processing, computer vision and  ... 
arXiv:1803.08680v4 fatcat:2daso6mi3fgdrb3jyvdr4w3o6y

Hybrid-DNNs: Hybrid Deep Neural Networks for Mixed Inputs [article]

Zhenyu Yuan, Yuxin Jiang, Jingjing Li, Handong Huang
2020 arXiv   pre-print
We develop a general architecture of hybrid deep neural networks (HDNNs) to support mixed inputs.  ...  For feature learning, convolutional neural networks (CNN) and multilayer perceptron (MLP) network are configured to separately process structural and numerical inputs.  ...  Qiu for valuable discussions on production characterization of the adopted oil block.  ... 
arXiv:2005.08419v1 fatcat:jqi3jbsukve6laa365jzfexcea

Deep Defense: Training DNNs with Improved Adversarial Robustness [article]

Ziang Yan, Yiwen Guo, Changshui Zhang
2018 arXiv   pre-print
Despite the efficacy on a variety of computer vision tasks, deep neural networks (DNNs) are vulnerable to adversarial attacks, limiting their applications in security-critical systems.  ...  ., adversarial examples) to fool well-trained DNN classifiers into making arbitrary predictions. To address this problem, we propose a training recipe named "deep defense".  ...  In the main body of our paper, we mimic the DeepFool attack calculation using a neural network.  ... 
arXiv:1803.00404v3 fatcat:kqo5qp5zkfdrhmenqitkjkljv4

CheMixNet: Mixed DNN Architectures for Predicting Chemical Properties using Multiple Molecular Representations [article]

Arindam Paul, Dipendra Jha, Reda Al-Bahrani, Wei-keng Liao, Alok Choudhary, Ankit Agrawal
2018 arXiv   pre-print
In this work, we present CheMixNet -- a set of neural networks for predicting chemical properties from a mixture of features learned from the two molecular representations -- SMILES as sequences and molecular  ...  The proposed CheMixNet models not only outperforms the candidate neural architectures such as contemporary fully connected networks that uses molecular fingerprints and 1-D CNN and RNN models trained SMILES  ...  Department of Commerce, National Institute of Standards and Technology as part of the Center for Hierarchical Materials Design (CHiMaD).  ... 
arXiv:1811.08283v2 fatcat:zigst4l7szhf5dzophrqmcqt2m

Sparse DNNs with Improved Adversarial Robustness [article]

Yiwen Guo, Chao Zhang, Changshui Zhang, Yurong Chen
2019 arXiv   pre-print
Deep neural networks (DNNs) are computationally/memory-intensive and vulnerable to adversarial attacks, making them prohibitive in some real-world applications.  ...  Our analyses reveal, both theoretically and empirically, that nonlinear DNN-based classifiers behave differently under l_2 attacks from some linear ones.  ...  Acknowledgement We would like to thank anonymous reviewers for their constructive suggestions.  ... 
arXiv:1810.09619v2 fatcat:5hm3tyfoz5hg5edg27khnkjcym

Nonlinear Reduced DNN Models for State Estimation [article]

Wolfgang Dahmen, Min Wang, Zhu Wang
2022 arXiv   pre-print
Neural Networks.  ...  We propose in this paper a data driven state estimation scheme for generating nonlinear reduced models for parametric families of PDEs, directly providing data-to-state maps, represented in terms of Deep  ...  We thank the reviewers for their valuable comments regarding the presentation of the material.  ... 
arXiv:2110.08951v2 fatcat:2kxkco656beizhdpticiarervq

A DNN Framework For Text Image Rectification From Planar Transformations [article]

Chengzhe Yan, Jie Hu, Changshui Zhang
2016 arXiv   pre-print
We explored the capability of deep neural network in learning geometric transformation and found the model could segment the text image without explicit supervised segmentation information.  ...  In this paper, a novel neural network architecture is proposed attempting to rectify text images with mild assumptions.  ...  Recently, some researches try to solve affine and perspective transformation using deep neural network.  ... 
arXiv:1611.04298v1 fatcat:oddj6uu455efjcyaqehbghbixe

Cracking open the DNN black-box

John Emmons, Sadjad Fouladi, Ganesh Ananthanarayanan, Shivaram Venkataraman, Silvio Savarese, Keith Winstein
2019 Proceedings of the 2019 Workshop on Hot Topics in Video Analytics and Intelligent Edges - HotEdgeVideo'19  
Advancements in deep neural networks (DNNs) and widespread deployment of video cameras have fueled the need for video analytics systems.  ...  We present promising results from preliminary work in efficiently encoding the intermediate activations sent between layers of a neural network and describe opportunities for further research.  ...  The compressed HD version (right) is small enough to be processed on device or sent to the cloud, but leads to inaccurate reading of the license plate. in deep neural networks (DNNs) on a variety of computer  ... 
doi:10.1145/3349614.3356023 dblp:conf/mobicom/EmmonsFAVSW19 fatcat:z4psoigzdra4jlqtgskx5xbhga

Towards Scalable Uncertainty Aware DNN-based Wireless Localisation [article]

Artan Salihu, Stefan Schwarz, Markus Rupp
2021 arXiv   pre-print
Existing deep neural network (DNN) based wireless localization approaches typically do not capture uncertainty inherent in their estimates.  ...  In this work, we propose and evaluate variational and scalable DNN approaches to measure the uncertainty as a result of changing propagation conditions and the finite number of training samples.  ...  The financial support by the Austrian Federal Ministry for Digital and Economic Affairs and the National Foundation for Research, Technology and Development is gratefully acknowledged.  ... 
arXiv:2106.04697v1 fatcat:ttd3dnthlnhfvbz2qpd2vnhedy

Fused DNN: A deep neural network fusion approach to fast and robust pedestrian detection [article]

Xianzhi Du and Mostafa El-Khamy and Jungwon Lee and Larry S. Davis
2017 arXiv   pre-print
We propose a deep neural network fusion architecture for fast and robust pedestrian detection. The proposed network fusion architecture allows for parallel processing of multiple networks for speed.  ...  Next, multiple deep neural networks are used in parallel for further refinement of these pedestrian candidates.  ...  We propose a deep neural network fusion architecture to address the pedestrian detection problem, called Fused Deep Neural Network (F-DNN).  ... 
arXiv:1610.03466v2 fatcat:ta4rylplsrbdhgmshqmndvte2m
« Previous Showing results 1 — 15 out of 8,478 results