Filters








4,158 Hits in 10.8 sec

Digital Elevation Model enhancement using Deep Learning [article]

Casey Handmer
2021 arXiv   pre-print
We demonstrate high fidelity enhancement of planetary digital elevation models (DEMs) using optical images and deep learning with convolutional neural networks.  ...  Deep learning-based photoclinometry robustly recovers features obscured by non-ideal lighting conditions. Method can be automated at global scale.  ...  This research was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration.  ... 
arXiv:2101.04812v1 fatcat:ndedyqelunhq7iexk6hhm6anbm

Preprocessing for Unintended Conducted Emissions Classification with ResNet

Gregory Sheets, Philip Bingham, Mark B. Adams, David Bolme, Scott L. Stewart
2021 Applied Sciences  
As a side-benefit, much was learned regarding the best preprocessing to use with the selected deep network for the UCE collected from these low power consumer devices obtained via current transformers.  ...  progression of preprocessing and a deep neural network (ResNet architecture) to classify UCE data obtained via current transformers.  ...  Acknowledgments: Thanks to Olivia Shafer of ORNL for help with her technical writing edits. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/app11198808 fatcat:2qi2oxyfnfdulkrw7nrvqfqxj4

A Spatiotemporal Fusion Method Based on Multiscale Feature Extraction and Spatial Channel Attention Mechanism

Dajiang Lei, Gangsheng Ran, Liping Zhang, Weisheng Li
2022 Remote Sensing  
In recent years, spatiotemporal image fusion based on deep learning has received wide attention.  ...  Remote sensing satellite images with a high spatial and temporal resolution play a crucial role in Earth science applications.  ...  Deeper networks can learn richer feature information and correlation mappings, while deep residual learning for image recognition (ResNet) [36] makes it easier to train deep networks.  ... 
doi:10.3390/rs14030461 fatcat:5s36qvzxireibmufythw5ysazm

Quantifying the Chaos Level of Infants' Environment via Unsupervised Learning [article]

Priyanka Khante and Mai Lee Chang and Domingo Martinez, Kaya de Barbaro, Edison Thomaz
2019 arXiv   pre-print
These unsupervised techniques include hierarchical clustering using K-Means, clustering using self-organizing map (SOM) and deep learning.  ...  In this work, we use three unsupervised machine learning techniques to quantify household chaos in infants' homes.  ...  With the deep learning approach, a common challenge is the interpretability of the extracted features.  ... 
arXiv:1912.04844v1 fatcat:ouzxxr23knh5tdx3vdgafwihxy

Music Genre Classification using Deep Learning

Sheeba Fathima
2021 International Journal for Research in Applied Science and Engineering Technology  
In this paper, we propose two methods for boosting music genre classification with convolutional neural networks: 1) using a process inspired by residual learning to combine peak- and average pooling to  ...  Machine learning techniques were used to classify music genres in this research. Deep neural networks (DNN) have recently been demonstrated to be effective in a variety of classification tasks.  ...  is fed into a classification deep neural network.  ... 
doi:10.22214/ijraset.2021.36087 fatcat:2tkneqaehrb6nd6atng3ifgziq

Learning Representations from EEG with Deep Recurrent-Convolutional Neural Networks [article]

Pouya Bashivan, Irina Rish, Mohammed Yeasin, Noel Codella
2016 arXiv   pre-print
Next, we train a deep recurrent-convolutional network inspired by state-of-the-art video classification to learn robust representations from the sequence of images.  ...  The proposed approach is designed to preserve the spatial, spectral, and temporal structure of EEG which leads to finding features that are less sensitive to variations and distortions within each dimension  ...  Our approach is fundamentally different from the previous attempts to learn high-level representations from EEG using deep neural networks.  ... 
arXiv:1511.06448v3 fatcat:7izufqjvpzhalnvus42owkyobm

Seismic features and automatic discrimination of deep and shallow induced-microearthquakes using neural network and logistic regression

S. Mostafa Mousavi, Stephen P. Horton, Charles A. Langston, Borhan Samei
2016 Geophysical Journal International  
The used machine-learning techniques have application for efficient automatic classification of low energy signals recorded at one or more seismic stations.  ...  using logistic regression and artificial neural network models, respectively.  ...  To take a closer look at the seismic features and their correlations, we applied the X-means method on the first 17 features with highest merits presented in Table 5 .  ... 
doi:10.1093/gji/ggw258 fatcat:bmfodjkjz5hmtchzbu57mcawhi

Deep Learning Techniques for Speech Emotion Recognition, from Databases to Models

Babak Joze Abbaschian, Daniel Sierra-Sosa, Adel Elmaghraby
2021 Sensors  
The current study reviews deep learning approaches for SER with available datasets, followed by conventional machine learning techniques for speech emotion recognition.  ...  Ultimately, we present a multi-aspect comparison between practical neural network approaches in speech emotion recognition.  ...  convolutional and recurrent neural networks as a deep learning method.  ... 
doi:10.3390/s21041249 pmid:33578714 pmcid:PMC7916477 fatcat:nj5ihjhvnfcxtk7hu3n4zx4bka

LC-RNN: A Deep Learning Model for Traffic Speed Prediction

Zhongjian Lv, Jiajie Xu, Kai Zheng, Hongzhi Yin, Pengpeng Zhao, Xiaofang Zhou
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
Furthermore, since traffic evolution is restricted by the underlying road network, a network embedded convolution structure is proposed to capture topology aware features.  ...  The fusion with other information, including periodicity and context factors, is also considered to further improve accuracy.  ...  ., 2016] is a spectral approach that ensures strictly localized filter and low computational complexity as well.  ... 
doi:10.24963/ijcai.2018/482 dblp:conf/ijcai/LvX0YZZ18 fatcat:ufu6ezwjkzhjlcdcegqbtaesiq

CRED: A Deep Residual Network of Convolutional and Recurrent Units for Earthquake Signal Detection [article]

S. Mostafa Mousavi, Weiqiang Zhu, Yixiao Sheng, Gregory C. Beroza
2018 arXiv   pre-print
Here, we introduce the Cnn-Rnn Earthquake Detector (CRED), a detector based on deep neural networks.  ...  The robustness of the trained model with respect to the noise level and non-earthquake signals is shown by applying it to a set of semi-synthetic signals.  ...  A closer look at a standard RNN unit in the previous figure.  ... 
arXiv:1810.01965v1 fatcat:tm33346ewrgyppgtmrco4e6ppi

Deep neural networks for the evaluation and design of photonic devices [article]

Jiaqi Jiang, Mingkun Chen, Jonathan A. Fan
2020 arXiv   pre-print
We will also examine how deep generative networks can learn geometric features in device distributions and even be configured to serve as robust global optimizers.  ...  In this Review, we will show how deep neural networks, configured as discriminative networks, can learn from training sets and operate as high-speed surrogate electromagnetic solvers.  ...  With layer-by-layer processing of data inputted to the network, data features with higher levels of abstraction are captured from lower level features, and complex network input-output relations can be  ... 
arXiv:2007.00084v1 fatcat:s76z7d6ghfdd7c3pe6f5guhksa

Deep convolutional embedding for digitized painting clustering [article]

Giovanna Castellano, Gennaro Vessio
2020 arXiv   pre-print
optimized with the task of finding a set of cluster centroids in this latent feature space.  ...  To address these issues, we propose to use a deep convolutional embedding model for digitized painting clustering, in which the task of mapping the raw input data to an abstract, latent space is jointly  ...  the low-level colour and texture features.  ... 
arXiv:2003.08597v2 fatcat:mhxomrao65g5vmeuwx3ehp3lsu

Bottom-up Attention, Models of [article]

Ali Borji, Hamed R. Tavakoli, Zoya Bylinskii
2019 arXiv   pre-print
In spite of tremendous efforts and huge progress, there is still room for improvement in terms finer-grained analysis of deep saliency models, evaluation measures, datasets, annotation methods, cognitive  ...  Sometimes deep models neglect low-level image features (local contrast) and overweight the contribution of high-level features (e.g. faces or text).  ...  Early computational saliency models were primarily concerned with identifying conspicuous regions due to low-level feature contrast (e.g. pop-out search).  ... 
arXiv:1810.05680v3 fatcat:yurrgypswnbt5kyxnxugivdwiq

InverseNet: Solving Inverse Problems with Splitting Networks [article]

Kai Fan, Qi Wei, Wenlin Wang, Amit Chakraborty, Katherine Heller
2017 arXiv   pre-print
We propose a new method that uses deep learning techniques to solve the inverse problems.  ...  The two networks are trained jointly to learn the end-to-end mapping, getting rid of a two-step training.  ...  Similar to the strategy in [17, 40] , we design a deep convolutional neural network to learn the inversion.  ... 
arXiv:1712.00202v1 fatcat:p3aootffcjgevmcqp27vlgxf3y

Research on super-resolution reconstruction of remote sensing images: a comprehensive review

Hui Liu, Yurong Qian, Xiwu Zhong, Long Chen, Guangqi Yang
2021 Optical Engineering: The Journal of SPIE  
, and deep-learning-based methods.  ...  The experimental results indicate the advantages and limitations of single-and multi-frame reconstruction, with the latter showing a higher performance.  ...  The network uses residuals and skip connections in a very deep architecture to transmit information processed at different abstraction levels and alleviate data degradation.  ... 
doi:10.1117/1.oe.60.10.100901 fatcat:44ssaq55ebfkrghyatvo5lbr2m
« Previous Showing results 1 — 15 out of 4,158 results