Who is the winner? Memristive-CMOS hybrid modules: CNN-LSTM versus HTM.

07:00 EST 28th November 2019 | BioPortfolio

Summary of "Who is the winner? Memristive-CMOS hybrid modules: CNN-LSTM versus HTM."

Hierarchical, modular and sparse information processing are signature characteristics of biological neural networks. These aspects have been the backbone of several artificial neural network designs of the brain-like networks, including Hierarchical Temporal Memory (HTM). The main contribution of this work is showing that Convolutional Neural Network (CNN) in combination with Long short term memory (LSTM) can be a good alternative for implementing the hierarchy, modularity and sparsity of information processing. To demonstrate this, we draw a comparison of CNN-LSTM and HTM performance on a face recognition problem with a small training set. We also present the analog CMOS-memristor circuit blocks required to implement such a scheme. The presented memristive implementations of the CNN-LSTM architecture are easier to implement, train and offer higher recognition performance than the HTM. The study also includes memristor variability and failure analysis.


Journal Details

This article was published in the following journal.

Name: IEEE transactions on biomedical circuits and systems
ISSN: 1940-9990


DeepDyve research library

PubMed Articles [5796 Associated PubMed Articles listed on BioPortfolio]

Empirical mode decomposition based long short-term memory neural network forecasting model for the short-term metro passenger flow.

Short-term metro passenger flow forecasting is an essential component of intelligent transportation systems (ITS) and can be applied to optimize the passenger flow organization of a station and offer ...

Hybrid LAE-CMOS Force-sensing System Employing TFT-based Compressed Sensing for Scalability of Tactile Sensing Skins.

Tactile sensing requires form-fitting and dense sensor arrays over large-areas. Hybrid systems, combining Large-Area Electronics (LAE) and silicon-CMOS ICs to respectively provide diverse sensing and ...

On the validity of memristor modeling in the neural network literature.

An analysis of the literature shows that there are two types of non-memristive models that have been widely used in the modeling of so-called "memristive" neural networks. Here, we demonstrate that su...

Stacked Bidirectional Convolutional LSTMs for Deriving 3D Non-contrast CT from Spatiotemporal 4D CT.

The imaging workup in acute stroke can be simplified by deriving non-contrast CT (NCCT) from CT perfusion (CTP) images. This results in reduced workup time and radiation dose. To achieve this, we pres...

Prototyping a memristive-based device to analyze neuronal excitability.

Many efforts have been spent in the last decade for the development of nanoscale synaptic devices integrated into neuromorphic circuits, trying to emulate the behavior of natural synapses. The study o...

Clinical Trials [2902 Associated Clinical Trials listed on BioPortfolio]

Hybrid SA RGP Center/S-H Skirt Daily Wear 90 Day Multicenter Study

The purpose of the study is to demonstrate that the SynergEyes SA Hybrid Contact Lens clinical performance is substantially equivalent to that of the SynergEyes A Hybrid Lens when studied:...

Concomitant Versus Hybrid Regimen for H. Pylori Eradication

10 day concomitant versus 14 day hybrid regimen as first line H. pylori eradication treatment in a high clarithromycin resistance area. A multicenter, randomized, equivalence trial.

Hybrid APC Assisted EMR for Large Colon Polyps

The purpose of this study to evaluate and examine, weather use of Hybrid Argon Plasma Coagulation (APC) as adjunct to endoscopic mucosal resection (EMR) will reduce the risk of residual or...

Comparison of One-stop Hybrid Revascularization Versus Percutaneous Coronary Intervention for the Treatment of Multi-vessel Disease

The "one-stop" hybrid coronary revascularization combines minimally invasive direct coronary artery bypass (MIDCAB) and PCI to be performed in the hybrid operating suite, an enhanced opera...

Prospective, Randomized Comparison of Hybrid Ablation vs. Catheter Ablation (PRHACA)

The overall objective of the study is to compare and assess the clinical outcomes of the hybrid procedure versus the catheter ablation alone in a randomized, prospective population for lon...

Medical and Biotech [MESH] Definitions

Screening techniques first developed in yeast to identify genes encoding interacting proteins. Variations are used to evaluate interplay between proteins and other molecules. Two-hybrid techniques refer to analysis for protein-protein interactions, one-hybrid for DNA-protein interactions, three-hybrid interactions for RNA-protein interactions or ligand-based interactions. Reverse n-hybrid techniques refer to analysis for mutations or other small molecules that dissociate known interactions.

Tailored macromolecules harboring covalently-bound biologically active modules that target specific tissues and cells. The active modules or functional groups can include drugs, prodrugs, antibodies, and oligonucleotides, which can act synergistically and be multitargeting.

A self-administered version of the Primary Care Evaluation of Mental Disorders (PRIME-MD), a diagnostic tool containing modules on multiple mental health disorders including anxiety, alcohol, eating, and somatoform modules. The Patient Health Questionnaire (PHQ-9) is designed specifically for mood/depression scoring each of the 9 DSM-IV criteria based on the mood module from the original PRIME-MD. The Generalized Anxiety Disorder scale (GAD-7) scores 7 common anxiety symptoms.

The procedures involved in combining separately developed modules, components, or subsystems so that they work together as a complete system. (From McGraw-Hill Dictionary of Scientific and Technical Terms, 4th ed)

The adaptive superiority of the heterozygous GENOTYPE with respect to one or more characters in comparison with the corresponding HOMOZYGOTE.

Quick Search

DeepDyve research library

Relevant Topic

Biological Therapy
Biological therapy involves the use of living organisms, substances derived from living organisms, or laboratory-produced versions of such substances to treat disease. Some biological therapies for cancer use vaccines or bacteria to stimulate the body&rs...

Searches Linking to this Article