Radiometric Resolution Analysis and a Simulation Model

Kaisti, Matti; Altti, Miikka; Poutanen, T.

doi:10.3390/rs8020085

Cited by 9 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3) Noise realizations: Time domain noise realization including white and power-law noise that corresponds a given power spectrum was generated [16], [17]. First a power spectral density (PSD) was defined as…”

Section: A Synthetic Electrocardiogram Generatormentioning

confidence: 99%

Training neural networks with synthetic electrocardiograms

Kaisti¹,

Laitala²,

Airola³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We present a method for training neural networks with synthetic electrocardiograms that mimic signals produced by a wearable single lead electrocardiogram monitor. We use domain randomization where the synthetic signal properties such as the waveform shape, RR-intervals and noise are varied for every training example. Models trained with synthetic data are compared to their counterparts trained with real data. Detection of r-waves in electrocardiograms recorded during different physical activities and in atrial fibrillation is used to compare the models. By allowing the randomization to increase beyond what is typically observed in the real-world data the performance is on par or superseding the performance of networks trained with real data. Experiments show robust performance with different seeds and training examples on different test sets without any test set specific tuning. The method makes possible to train neural networks using practically free-to-collect data with accurate labels without the need for manual annotations and it opens up the possibility of extending the use of synthetic data on cardiac disease classification when disease specific a priori information is used in the electrocardiogram generation. Additionally the distribution of data can be controlled eliminating class imbalances that are typically observed in health related data and additionally the generated data is inherently private.

show abstract

Section: A Synthetic Electrocardiogram Generatormentioning

confidence: 99%

Training neural networks with synthetic electrocardiograms

Kaisti¹,

Laitala²,

Airola³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The noises caused by the insertion loss of the microwave switch also are not included in the expression (1). This effect caused by the implementation of the differential measurement method and the equality of insertion loss between both directions of the microwave switch [22]. If the last condition is not met, then noise loss accounting is required.…”

Section: Measurement Errors Analysismentioning

confidence: 99%

Laboratory model of microwave hyperspectrometer for internal radiation researches of layered natural mediums

Ubaychin¹,

Суржиков²,

Stary³

et al. 2020

View full text Add to dashboard Cite

This paper describes an achieved result in designing of the four-receiver microwave hyperspectrometer based on zero measurement method and the multi-receiver concept of realization. The block diagrams of the microwave front-end with operating frequency band from 18 to 26.5 GHz, the radiometric receiver, and the frequency transmission module are shown. The technical implementation of every described blocks of the microwave hyperspectrometer are described. Presented technical implementation includes a list of modern components are used to designing of laboratory model of hyperspectrometer. The detailed block diagram of the microwave front-end is presented. Description of main technical characteristics of parts of the microwave front-end is also presented. The analysis of a technical characteristics influence to measurement errors is shown. The mathematical model of a measurement error calculation in hyperspectral modeis described. The influence of the active reference noise generator temperature accuracy, the directional coupler insertion loss and the passive reference noise generator temperature accuracy to a measurement errors are described. A numerical experimental results of designed mathematical model are presented. The dependence of the minimum value of measurement errors form the temperature of passive noise generator is presented.

show abstract

“…However, no significant techniques were found to improve the design of radiometric receivers. The radiometric receivers currently used are based on total power or modulation principle proposed by Dick's in 1946 [2]. It is wellknown, that the receiver gain fluctuation has a major impact on sensitivity of these receivers.…”

Section: Study Objectivesmentioning

confidence: 99%