JP2015022060A - Information providing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide information using characters and voices of the same language as the language used by a passerby and output various pieces of information in the language which the passerby can understand.SOLUTION: Speech signals of the conversations of passers-by input from a speech input part 11 are subjected to language recognition processing for specifying a language to be used using speech dictionary database units 14b, 14c by a speech recognition part 15. A recognition rate N of each language input by the speech recognition part 15 is calculated, and a language for which the recognition rate N is the highest is specified as the language spoken by the passers-by. The spoken langue is output from a display 12 and a speaker 13 by display information and speech information using the specified language.

Description

  The present invention relates to an information providing apparatus that is installed in, for example, a station premises, an airport lobby, and the like and provides information using characters and voices in the same language as a language used by a nearby passerby.

  In Patent Document 1, an LED display, a discharge system such as plasma, a CRT display, and a liquid crystal display are arranged in a station premises, an airport lobby, etc., providing political / public information from this image display, an emergency accident・ It is disclosed to provide information at the time of disaster, site guidance, and corporate / store advertising.

  As shown in FIG. 3 of Patent Document 1, characters to be displayed on the image display device are displayed in Japanese. In addition, in a station yard, an airport lobby, and the like, an information providing device that alternately displays multilingual languages such as Japanese and English is installed in order to provide various information to foreigners.

JP 2002-304574 A

  For example, there may be a large number of non-Japanese passengers in the arrival lobby when international planes arrive at the airport. A problem that many foreigners do not know the displayed content when providing information such as guidance routes and advertisement information in Japanese using an information providing device such as the image display device of Patent Document 1. There is.

  Further, even if an information providing apparatus that displays Japanese and English alternately or in parallel is used, the display contents of guidance information and advertisement information are not understood by foreigners who cannot understand both Japanese and English.

  The objective of this invention is providing the information provision apparatus which eliminates the above-mentioned subject and provides information with the language which a passerby nearby uses.

  In order to achieve the above object, an information providing apparatus according to claim 1 includes: a voice input unit; a voice identification unit that identifies which language is based on a voice signal input from the voice input unit; And a video output unit that outputs video information including characters in the language identified by the voice identification unit as provided information.

  According to a second aspect of the present invention, there is provided an information providing apparatus including a voice output unit that outputs voice information in a language identified by the voice identification unit as provided information.

  The information providing apparatus according to claim 3 is characterized in that the provided information includes at least guidance information or advertisement information.

  The information providing apparatus according to claim 4, wherein the speech identification unit includes a speech dictionary database unit of a plurality of languages, an analysis speech signal obtained by performing a matching process on the speech signal, a vocabulary for each speech dictionary database unit, and The recognition rate of words recognized for each language is calculated by comparing with pronunciation, and provided information in a language with the highest recognition rate is output from the output unit.

  The information providing apparatus according to claim 5 is characterized in that the video output unit displays the language having the highest recognition rate and the language having the next highest recognition rate alternately or in parallel.

  The information providing apparatus according to claim 6 is installed in a station premises or an airport.

  The information providing apparatus according to the present invention provides information using characters and voices in the same language as the language used by nearby passers-by, thereby enabling the passer-by to understand display information and voice information. Various kinds of information can be effectively provided.

It is a block circuit block diagram of the Example of an information provision apparatus. It is a flowchart figure of the process part which determines the language of an audio | voice. It is explanatory drawing of the example of a display output by each language.

The present invention will be described in detail based on the embodiments shown in the drawings.
FIG. 1 is a block circuit configuration diagram of the information providing apparatus 10. The information providing apparatus 10 includes an audio input unit 11 including a sound collecting microphone for inputting a passer's voice, a video output unit that displays various types of information such as route guidance as video information, and an audio output. A speaker unit 13 that provides various information according to display contents by voice, a voice dictionary database unit 14 that stores vocabulary and pronunciation of each language, a voice input unit 11, a display unit 12, and a speaker unit 13 and a voice identification unit 15 connected to the voice dictionary database unit 14. The speech dictionary database unit 14 includes a Japanese speech dictionary database unit 14a, an English speech dictionary database unit 14b, a Chinese speech dictionary database unit 14c,.

  Since the speech identification unit 15 places a burden on language identification processing, a plurality of speech identification units 15 may be provided in parallel so as to correspond to the speech dictionary database units 14a, 14b, 14c,. In addition, the voice identification unit 15 includes a storage unit, and stores in advance video information including characters in each language and voice information in each language.

  The voice input unit 11 mainly inputs a voice spoken by a passerby in the vicinity of the display unit 12 and the speaker unit 13, and the voice input unit 11 is preferably arranged on the upstream side of the passer's pass direction. By arranging in this way on the upstream side, the timing of providing information via the display unit 12 and the speaker unit 13 to the passer-by even if the voice identification unit 15 takes some time to process. Easy to match.

  The display unit 12 employs a liquid crystal display, a full-color LED display, or the like as appropriate, and the display unit 12 has an appropriate size according to the installation location such as a station premises, an airport, or a commercial facility. For example, when installed in an airport passage, a display screen consisting of 100 inches or more is desirable.

  Various types of provision information displayed on the display screen of the display unit 12 are guidance information, advertisement information, and the like for guiding destinations and routes, and are displayed using characters and images. When the information providing apparatus 10 is installed in a station or an airport, guidance information is mainly displayed. When the information providing apparatus 10 is installed in a commercial facility or the like, advertisement information is mainly displayed. The guidance information and the advertisement information may be alternately switched and displayed at predetermined time intervals. The speaker unit 13 outputs voice and music according to the contents of the guidance information and advertisement information.

  The voice identification unit 15 may be connected to the terminal device T installed in the management center or the like by wireless or wired. Update of the video information and audio information and the update of the dictionary database units 14a, 14b, 14c,... For each language can be performed from the terminal device T by wireless or wired. Emergency information such as an earthquake or fire may be received wirelessly or by wire, and information such as evacuation guidance may be provided by interrupting the current display output and audio content.

  FIG. 2 is a flowchart of the voice identification unit 15 that determines the language of the input voice. In step S1, a voice signal of a passer's conversation is input from the voice input unit 11, and in step S2, the input voice signal is A / D converted.

  In steps S3 to S5, the speech recognition unit 15 analyzes which language the A / D converted speech signal is, and performs language recognition processing. In step S3, the A / D converted speech signal is subjected to a matching process of an acoustic model that holds a phoneme substantially equivalent to one Roman letter or a syllable frequency pattern equivalent to one Kana character.

  This acoustic model needs to use a model according to the voice input environment. If the installation location of the information providing apparatus 10 is a noisy input environment, the acoustic model learned based on the voice recorded under the noise. Will be used.

  In step S4, the word that identifies the word corresponding to the analysis voice signal is compared with the analysis voice signal matched by the voice identification unit 15, the vocabulary of the voice dictionary database units 14a, 14b, 14c,. Perform identification processing. The speech dictionary database unit 14 may prepare a plurality of languages other than Japanese, English, and Chinese. Moreover, you may select the audio | voice dictionary database part 14 to use suitably from several audio | voice dictionary database parts 14a, 14b, 14c, ... according to a time zone and a day of the week.

  In step S5, the speech recognition unit 15 calculates a recognition rate N for each language for the word string identified by the word identification process, and identifies a language having a high recognition rate N. In the word string identified in step S4, there are some words that cannot be identified. The recognition rate N when the total number of words subjected to word identification processing is used as the denominator and the number of identified words is used as the numerator. calculate.

  If speech is input in Japanese via the speech input unit 11, the recognition rate N calculated using the Japanese acoustic model and the Japanese speech dictionary database unit 14a is the highest numerical value. If the recognition rate N is calculated by the acoustic models of other languages and the speech dictionary database units 14b, 14c,..., The words of other languages corresponding to the analysis speech signal based on Japanese pronunciation cannot be identified. Low value.

  The recognition rate N is calculated from the processing in steps S3 and S4 for each language of the speech of the passerby input in this way, and the language in which the passerby speaks the language having the highest recognition rate N in step S5. As specified. In addition to the recognition rate N, this identification process can be performed with the number of recognized words M. This is because the number of recognizable words M is the largest when the language recognition processing is performed on the language in conversation by the acoustic model and the speech dictionary database unit 14 of the same language.

  In step S6, using the specified language, the display unit 12 and the speaker unit 13 output the video information including the characters of the specific language stored in advance and the audio information of the specific language.

  FIG. 3 shows a display example of the display unit 12 output and displayed in each language when the language is specified. For example, when providing information on the destination direction of the baggage claim area for passengers on an international flight that has arrived, if it is specified that the passerby mainly speaks Japanese, the display unit 12 installed in the aisle The display contents as shown in (a) are selectively displayed. In addition, when it is specified that the passerby mainly speaks English, the display content as shown in (b) is selectively displayed. Further, when it is specified that the passerby mainly speaks Chinese, the display contents as shown in (c) are selected and displayed. It is preferable to perform guidance and sound output from the speaker unit 13 in the languages corresponding to these displays shown in FIGS.

  It should be noted that once the language is specified and the provision of information is started, the language is continuously output for a predetermined time. This is because confusion arises when switching to a different language output each time a language is specified.

  Further, a plurality of display units 12 may be installed, and a plurality of types of languages may be displayed separately in descending order of recognition rate N. A plurality of languages having a high recognition rate N may be displayed alternately or in parallel on the display screen of one display unit 12. However, the output from the speaker unit 13 is output by selecting one of the languages.

DESCRIPTION OF SYMBOLS 10 Information provision apparatus 11 Voice input part 12 Display part 13 Speaker part 14a, 14b, 14c Voice dictionary database part 15 Voice identification part

Claims (6)

  Provided information is a voice input unit, a voice identification unit that identifies which language is based on a voice signal input from the voice input unit, and video information including characters in the language identified by the voice identification unit An information providing apparatus comprising a video output unit for outputting.   The information providing apparatus according to claim 1, further comprising a voice output unit that outputs voice information in a language identified by the voice identification unit as provided information.   The information providing apparatus according to claim 1, wherein the provision information includes at least guidance information or advertisement information.   The speech identification unit includes a speech dictionary database unit of a plurality of languages, and compares an analysis speech signal that has been subjected to matching processing on the speech signal, a vocabulary and a pronunciation of each speech dictionary database unit, and recognizes each language The information providing apparatus according to any one of claims 1 to 3, wherein a recognition rate of the completed word is calculated, and provided information in a language having the highest recognition rate is output from the output unit.   The information providing apparatus according to claim 4, wherein the video output unit displays the language having the highest recognition rate and the language having the next highest recognition rate alternately or in parallel.   The information providing apparatus according to claim 1, wherein the information providing apparatus is installed in a station premises or in an airport.

Images