Abstract: A position estimator outputs, by inputting information obtained from the frame images into a trained model, positional information of a plurality of parts of a human body from the trained model. A determination processor outputs information on erroneous detection by the position estimator. The position estimator outputs a first position group in response to an input of an obtained first frame image, the first frame image being obtained by the receiver and outputs a second position group in response to an input of an obtained second frame image different from the first frame image, the second frame image being obtained by the receiver. The determination processor outputs the information on erroneous detection by the position estimator based on a degree of difference between the first position group and the second position group and a degree of matching between the first frame image and the second frame image.

Abstract: A sleep state of a subject is estimated from respiration signals that are readily detected without restraining the subject. From a voltage waveform based on respiratory movement of a person, positive peak values and the (time) interval between adjacent peaks are calculated. Also obtained is the area of a region that is enclosed by the voltage waveform and a time axis between peaks, and the average value and dispersion value of the obtained inter-peak areas are calculated. The calculated values are used to set a preliminary sleep state value for a given period, and sleep state values over plural periods are factored in to estimate a sleep state at a given time point.

Abstract: An object of the present invention is to provide a sleep state estimation device capable of estimating a sleep state (deep sleep, light sleep) in real time with relative simplicity. Voltage variations resulting from human respiratory movement are measured for each given period to calculate from the measurement results positive voltage peak values and time intervals between adjacent peaks. The obtained peak values and peak intervals are used to calculate a mean peak interval value A, a coefficient of variation B based on dispersion of the peak intervals, and a coefficient of variation C based on the peak values. The sleep state in this period is estimated from the calculation results. More specifically, referring to FIG.

Abstract: A sleep state of a subject is estimated from respiration signals that are readily detected without restraining the subject. From a voltage waveform based on respiratory movement of a person, positive peak values and the (time) interval between adjacent peaks are calculated. Also obtained is the area of a region that is enclosed by the voltage waveform and a time axis between peaks, and the average value and dispersion value of the obtained inter-peak areas are calculated. The calculated values are used to set a preliminary sleep state value for a given period, and sleep state values over plural periods are factored in to estimate a sleep state at a given time point.

Abstract: An object of the present invention is to provide a sleep state estimation device capable of estimating a sleep state (deep sleep, light sleep) in real time with relative simplicity. Voltage variations resulting from human respiratory movement are measured for each given period to calculate from the measurement results positive voltage peak values and time intervals between adjacent peaks. The obtained peak values and peak intervals are used to calculate a mean peak interval value A, a coefficient of variation B based on dispersion of the peak intervals, and a coefficient of variation C based on the peak values. The sleep state in this period is estimated from the calculation results. More specifically, referring to FIG.

Abstract: The present invention provides a new sleep analysis technique with which sleep analysis with relatively high reliability can be simply and speedily performed by using sleeping depth data. An appearance ratio conversion processing portion reads sleeping depth data (qualitative variable) stored in a RAM to calculate a movement appearance ratio n-M.A.R. (quantitative variable) meeting a relation of sleeping depth=2 (deep sleep: SWS). Then, an evaluation result calculation processing portion reads the calculated movement appearance ratio data (quantitative variable) from the RAM, and processes the calculated movement appearance ratio data in accordance with an evaluation rule in an evaluation rule storage portion to calculate an evaluation score for the evaluation factor concerned. When the evaluation factor is “rapidity of falling asleep” for example, a time period from a point when a person goes to bed to a point when the movement appearance ratio n-M.A.R.