International Business Machines Corporation

Yasunori Yamada

Kaoru Shinkawa

Kun Zhao

Masatomo Kobayashi

Tetsuro Morimura

Takayuki Katsuki

Satoshi Hara

A method, computer system, and a computer program product for speech feature distribution is provided. The present invention may include receiving two or more speech samples from a user. The present invention may include dividing the two or more speech samples into a a plurality of pause segments. The present invention may include determining a pause duration for each of the plurality of pause segments. The present invention may include determining a distribution of pause durations. The present invention may include determining a distance between the distribution of pause durations.

Detection of cognitive impairment using speech feature distribution

Aspects of the present disclosure relate to a system receiving an infrared (IR) image, receiving a depth map associated with the IR image, applying one or more filters to the IR image based on a set of parameters, generating a 2D pose for a subject based on the filtered IR image, and generating a 3D pose by correlating the 2D pose to the depth map.

Filtered images for 3D body tracking

Optimized learning settings of neural networks are efficiently determined by an apparatus including a processor and one or more computer readable mediums collectively including instructions that, when executed by the processor, cause the processor to train a first neural network with a learning setting; extract tentative weight data from the first neural network with the learning setting; calculate an evaluation value of the first neural network with the learning setting; and generate a predictive model for predicting an evaluation value of a second neural network with a new setting based on tentative weight data of the second neural network by using a relationship between the tentative weight data of the first neural network and the evaluation value of the first neural network.

Efficient determination of optimized learning settings of neural networks

A computer-implemented method for supporting detection of a sign of cognitive decline is disclosed. In the method, two or more sets of conversation data of a target individual are obtained. The method includes combining at least two sets of the conversation data corresponding to plural different days to generate a combined set and extracting one or more linguistic features that represent vocabulary richness from at least the combined set. A feature that characterizes word repetitiveness in conversations across the plural different days of the target individual is calculated from the one or more linguistic features. The feature that characterizes the word repetitiveness is output.

Word repetition in separate conversations for detecting a sign of cognitive decline

A technique for estimating activity of a plurality of muscles is disclosed. In this technique, for each of two or more muscles included in a musculoskeletal body model, a set of time-series data of muscle state change that accompanies motion of a body is acquired. A correlation between the sets of time-series data for a pair of muscles included among the two or more muscles is calculated. An activity of a plurality of muscles included in the musculoskeletal body model in a motion state that is an estimation target is estimated using an objective function including a constraint using the correlation of the sets of time-series data.

Estimation of muscle activities using the muscles relationship during simulating movements

A computer-implemented method for estimating an age of a target individual includes obtaining information of eye movement of the target individual, analyzing the information of the eye movement to extract a feature of the eye movement, and estimating the age of the target individual using the feature of the eye movement.

Age estimation using feature of eye movement

Volume rendering is performed by a method, comprising: obtaining an original volume data, transforming the original volume data based on a distance from a viewpoint to the original volume data, to generate transformed volume data, generating particles from the transformed volume data, and projecting the particles on an image plane to obtain a 2D image corresponding to the original volume data.

Growing community of inventors

Saitama, Japan

Yasunori Yamada