Hitachi, Ltd.

Maxell, Ltd.

National Institute of Information and Communications Technology, Incorporated

Johns Hopkins University

Naoyuki Kanda

Takashi Sumiyoshi

Yusuke Fujita

Yo Nonomura

Shota Horiguchi

Ryu Takeda

Toshiyuki Ono

Tatsuya Tokunaga

Yasunari Obuchi

Masao Ishiguro

Kenji Nagamatsu

Kiyoshi Kumagai

Chisa Nagai

Yo Takeuchi

Kenji Ohya

Shuhei Furuya

Shinji Watanabe

In a mixed reality display system, a server and a plurality of mixed reality display terminals are connected, and virtual objects are displayed. The virtual objects include a shared virtual object for which a plurality of terminals have an operation authority and a private virtual object for which only a specific terminal has an operation authority. The server has virtual object attribute information for displaying the virtual objects in each terminal, and each terminal has a motion detecting unit that detects a motion of a user for switching between the shared virtual object and the private virtual object. When a detection result by the motion detecting unit is received from the terminal, the server updates the virtual object attribute information depending on whether the virtual object is the shared virtual object or the private virtual object, and transmits data of the virtual object after the update to each terminal.

Mixed reality display system and mixed reality display terminal

A head-mounted display (HMD), which is operated by a gesture operation performed by a user, is provided with a distance image acquisition unitthat detects a gesture operation, a position information acquisition unitthat acquires position information of the HMD, and a communication unitthat performs communication with another HMD'. A control unitsets and displays an operating spacewhere a gesture operation performed by the useris valid, exchanges position information and operating space information of the host HMDand the other HMDtherebetween by the communication unit, and adjusts the operating space of the host HMD so that the operating spaceand an operating space′ of the other HMDdo not overlap each other.

Head-mounted display

An audio analysis platform may receive a portion of an audio input, wherein the audio input corresponds to audio associated with a plurality of speakers. The audio analysis platform may process, using a neural network, the portion of the audio input to determine voice activity of the plurality of speakers during the portion of the audio input, wherein the neural network is trained using reference audio data and reference diarization data corresponding to the reference audio data. The audio analysis platform may determine, based on the neural network being used to process the portion of the audio input, a diarization output associated with the portion of the audio input, wherein the diarization output indicates individual voice activity of the plurality of speakers. The audio analysis platform may provide the diarization output to indicate the individual voice activity of the plurality of speakers during the portion of the audio input.

Multi-speaker diarization of audio input using a neural network

In order to assist participants in thinking of an idea by acquiring audio data, it is provided a workshop assistance system, which includes a computer having an arithmetic apparatus configured to execute predetermined processing, a storage device coupled to the arithmetic apparatus, and a communication interface coupled to the arithmetic apparatus, the computer being configured to access solved problem case data including information of solved cases that correspond to problem data, the workshop assistance system comprising: a problem processing module configured to search, by the arithmetic apparatus, solved cases based on problem data that is generated from a discussion among participants; and an idea generation module configured to present, by the arithmetic apparatus, idea data including the generated problem data and information of the solved case found in the search to the participants.

Workshop assistance system and workshop assistance method

A head-mounted display (HMD), which is operated by a gesture operation performed by a user, is provided with a distance image acquisition unitthat detects a gesture operation, a position information acquisition unitthat acquires position information of the HMD, and a communication unitthat performs communication with another HMD'. A control unitsets and displays an operating spacewhere a gesture operation performed by the useris valid, exchanges position information and operating space information of the host HMDand the other HMDtherebetween by the communication unit, and adjusts the operating space of the host HMD so that the operating spaceand an operating spaceof the other HMDdo not overlap each other.

A speaker estimation method that estimate the speaker from audio and image includes: inputting audio; extracting a feature quantity representing a voice characteristic from the input audio; inputting an image; detecting person regions of respective persons from the input image; estimating feature quantities representing voice characteristics from the respective detected person regions; Performing a change such that an image taken from another position and with another angle is input when any person is not detected; calculating a similarity between the feature quantity representing the voice characteristic extracted from the audio and the feature quantity representing the voice characteristic estimated from the person region in the image; and estimating a speaker from the calculated similarity.

Speaker estimation method and speaker estimation device

A speech recognition device includes: an acoustic model based on an End-to-End neural network responsive to an observed sequence formed of prescribed acoustic features obtained from a speech signal by feature extracting unit, for calculating probability of the observed sequence being a certain symbol sequence; and a decoder responsive to a symbol sequence candidate, for decoding a speech signal by a WFST based on a posterior probability of each of word sequences corresponding to the symbol sequence candidate, probabilities calculated by the acoustic model for symbol sequences selected based on an observed sequence, and a posterior probability of each of the plurality of symbol sequences.

Speech recognition device and computer program

An object is to provide a speech recognition device with improved recognition accuracy using characteristics of a neural network. A speech recognition device includes: an acoustic modelimplemented by a RNN (recurrent neural network) for calculating, for each state sequence, the 45 posterior probability of a state sequence in response to an observed sequence consisting of prescribed speech features obtained from a speech; a WFSTbased on S-1HCLG calculating, for each word sequence, posterior probability of a word sequence in response to a state sequence; and a hypothesis selecting unit, performing speech recognition of the speech signal based on a score calculated for each hypothesis of aword sequence corresponding to the speech signal, using the posterior probabilities calculated by the acoustic modeland the WFSTfor the input observed sequence.

Provided is a voice search technology that can efficiently find and check a problematic call. To this end, a voice search system of the present invention includes a call search database that stores, for each of a reception channel and a transmission channel of each of a plurality of pieces of recorded call voice data, voice section sequences in association with predetermined keywords and time information. The call search database is searched based on an input search keyword, so that a voice section sequence that contains the search keyword is obtained. More specifically, the voice search system obtains, as a keyword search result, a voice section sequence that contains the search keyword and the appearance time thereof from the plurality of pieces of recorded call voice data, and obtains, based on the appearance time in the keyword search result, the start time of a voice section sequence of another channel immediately before the voice section sequence obtained as the keyword search result, and thus determines the start time as the playback start position for playing back the recorded voice. Then, the playback start position is output as a voice search result.

Voice search system, voice search method, and computer-readable storage medium

[Object] An object is to provide a training method of improving training of a recurrent neural network (RNN) using time-sequential data. [Solution] The training method includes a stepof initializing the RNN, and a training stepof training the RNN by designating a certain vector as a start position and optimizing various parameters to minimize error function. The training stepincludes: an updating stepof updating RNN parameters through Truncated BPTT using consecutive N (N&#x2265;3) vectors having a designated vector as a start point and using a reference value of a tail vector as a correct label; and a first repetition stepof repeating the process of executing the training step by newly designating a vector at a position satisfying a prescribed relation with the tail of N vectors used at the updating step until an end condition is satisfied. The vector at a position satisfying the prescribed relation is positioned at least two vectors behind the designated vector.

Recurrent neural network training method, computer program therefor and speech recognition device

The present invention pertains to a method for estimating a sound source position in a space with high accuracy using a microphone installed on a robot moving in the space. A mobile robot includes a self-position estimation unit configure to estimate the self-position of the mobile robot, a sound source information obtaining unit configured to obtain direction information of an observed sound source, and a sound source position estimation unit configured to estimate the position of the sound source based on the estimated self-position and the direction information of the sound source.

Growing community of inventors

Tokyo, Japan

Naoyuki Kanda