Results 71 to 80 of about 1,314,055 (250)

Efficient Super Resolution Using Binarized Neural Network [PDF]

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2019
Deep convolutional neural networks (DCNNs) have recently demonstrated high-quality results in single-image super-resolution (SR). DCNNs often suffer from over-parametrization and large amounts of redundancy, which results in inefficient inference and high memory usage, preventing massive applications on mobile devices.
Ma, Yinglan, Xiong, Hongyu, Hu, Zhe, Ma, Lizhuang   +3 more
openaire   +2 more sources

Binarized Simplicial Convolutional Neural Networks

Neural Networks
Graph Neural Networks have a limitation of solely processing features on graph nodes, neglecting data on high-dimensional structures such as edges and triangles. Simplicial Convolutional Neural Networks (SCNN) represent higher-order structures using simplicial complexes to break this limitation albeit still lacking time efficiency.
Yi Yan, Ercan Engin Kuruoglu
openaire   +3 more sources

Training Multi-Bit Quantized and Binarized Networks with a Learnable Symmetric Quantizer

IEEE Access, 2021
Quantizing weights and activations of deep neural networks is essential for deploying them in resource-constrained devices, or cloud platforms for at-scale services.
Phuoc Pham, Jacob A. Abraham, Jaeyong Chung   +2 more
doaj   +1 more source

On better training the infinite restricted Boltzmann machines

, 2017
The infinite restricted Boltzmann machine (iRBM) is an extension of the classic RBM. It enjoys a good property of automatically deciding the size of the hidden layer according to specific training data.
Gao, Xunzhang, Li, Xiang, Peng, Xuan
core   +1 more source

Structure–Function Decoupling of the Sensorimotor and Default Mode Networks in Black Americans With MS

Annals of Clinical and Translational Neurology, EarlyView.
ABSTRACT Background and Objectives Multiple sclerosis (MS) exhibits racially disparate rates of disease progression. Black people with MS (B‐PwMS) experience a more severe disease course than non‐Hispanic White people with MS (NHW‐PwMS). Here we investigated structural and functional connectivity as well as structure–function decoupling in the ...
Emilio Cipriano, Giacomo Boffa, Maria Petracca, Marta Ponzano, Nicole Graziano, Claire Wigley, Claire Riley, Jonathan Howard, Pietro Bontempi, Sylvia Klineova, Fred Lublin, Matilde Inglese   +11 more
wiley   +1 more source

A Deep Learning Accelerator Based on a Streaming Architecture for Binary Neural Networks

IEEE Access, 2022
Deep neural networks (DNNs) have played an increasingly important role in various areas such as computer vision and voice recognition. While training and validation become gradually feasible with high-end general-purpose processors such as graphical ...
Quang Hieu Vo, Ngoc Linh Le, Faaiz Asim, Lok-Won Kim, Choong Seon Hong   +4 more
doaj   +1 more source

Highly Tunable and Cell‐Remodelable Thiol‐ene Alginate‐Peptide Crosslinked Hydrogels to Recreate Cellular and Organoid Microenvironments for Biofabrication

Advanced Healthcare Materials, EarlyView.
Norbornene‐functionalized alginate is crosslinked with a di‐thiolated peptide sequence cleavable by cell‐secreted matrix metalloproteinases and decorated with cell‐adhesion peptides upon exposure to UV. Thyroid follicles, endometrial and intestinal organoids are encapsulated in a gentle manner with high cell viability and correct phenotype ...
Julia Fernández‐Pérez, Adrián Seijas‐Gamardo, Antonio J. Feliciano, Panagiota Kakni, Anna M. Kip, Paul Wieringa, Stefan Giselbrecht, Matthew B. Baker, Lorenzo Moroni   +8 more
wiley   +1 more source

SELECTION OF LINEAR AND ELLIPTICAL OBJECTS IN THE IMAGE BASED ON THE CALCULATION OF THE CORRELATION COEFFICIENT

Современная наука и инновации, 2022
The article presents the method of classification of the simplest geometric images based on the calculation of the correlation coefficient of connected objects in a binarized image.
V. V. Bulatov
doaj   +1 more source

Bird Sound Detection with Binarized Neural Networks

ELEKTRIKA- Journal of Electrical Engineering, 2022
By analysing the behavioural patterns of bird species in a specific region, researchers can predict future changes in the ecosystem. Many birds can be identified by their sounds, and autonomous recording units (ARUs) can capture real-time bird vocalisations. The recordings are analysed to see if there are any bird sounds.
Ahmad Zabidi, Muhammad Mun’im, Wong, Kah Liang, Sheikh, Usman Ullah, Sadiah, Shahidatul, Nurudin, Muhammad Afiq   +4 more
openaire   +2 more sources

Inverse‐Designed on‐Chip Terahertz Three‐Channel Mode and Wavelength Division Demultiplexer

Advanced Photonics Research, EarlyView.
An ultracompact terahertz demultiplexer is developed using a hybrid inverse design that combines genetic algorithms with topology optimization. The device enables simultaneous mode‐ and wavelength‐division multiplexing within a footprint smaller than two wavelengths.
Faqian Chong, Yulun Wu, Bingtao Gao, Shilong Li, Hongsheng Chen, Song Han   +5 more
wiley   +1 more source