Utterance recognition device, program, and utterance recognition method
The speech recognition device and method address the issue of inconsistent utterances in restricted areas by identifying and alerting to contradictions, ensuring accurate and safe access control.
Patent Information
- Application Number
- JP2024078168
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-05-13
- Publication Date
- 2025-11-26
AI Technical Summary
Existing technologies lack the ability to determine whether recognized route information is correct, particularly in access-restricted areas, leading to potential inconsistencies and safety risks.
A speech recognition device and method that acquire, recognize, and determine the content of utterances related to entry into restricted areas, identifying contradictions between administrator and respondent speech, and output alerts when inconsistencies are detected.
Ensures accurate determination of consistent utterances in restricted areas, preventing potential safety hazards by alerting to inconsistencies, thereby enhancing safety and efficiency in access control.
Smart Images

Figure 2025172582000001_ABST
Abstract
Description
[Technical Field]
[0001] The present disclosure relates to a speech recognition device, a program, and a speech recognition method. [Background technology]
[0002] Patent Document 1 describes that route information is automatically constructed using route-related voice control instructions from an air traffic controller. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-43200 Summary of the Invention [Problem to be solved by the invention]
[0004] However, there is no disclosure of how to determine whether the recognized route information is correct. Therefore, one of the objects of the present disclosure is to provide an utterance recognition device, a program, and an utterance recognition method that determine whether the content of a recognized utterance regarding entry into an entry-restricted area is inconsistent, and output a message indicating that an inconsistency is found. [Means for solving the problem]
[0005] The speech recognition device of the present disclosure comprises: a speech acquisition means for acquiring speech from a manager who manages an access-restricted area, which is an area with access restriction, or a respondent who responds to the manager; a recognition means for recognizing content of the acquired utterance related to entry into the access restriction area; a determination means for determining whether or not the recognized utterance has a contradiction in content related to entry into the entry-restricted area; The speech recognition device includes an output means for outputting a contradiction when the determination means determines that a contradiction exists.
[0006] The program of the present disclosure is Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; This is a program that causes the speech recognition device to output a contradiction when it is determined that a contradiction exists.
[0007] The speech recognition method of the present disclosure includes: The speech recognition device Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; This is a speech recognition method that outputs a contradiction when it is determined that a contradiction exists. [Effects of the Invention]
[0008] The present disclosure provides a speech recognition device, a program, and a speech recognition method that determine whether or not the content of a recognized utterance regarding entry into an entry-restricted area is inconsistent, and output a notice to that effect if there is an inconsistency. [Brief explanation of the drawings]
[0009] [Figure 1] 1 is a block diagram illustrating a configuration of an utterance recognition device according to the present disclosure. [Figure 2] 1 is a flowchart of a speech recognition method according to the present disclosure. [Figure 3] FIG. 1 illustrates an exemplary schematic of speech recognition according to the present disclosure. [Figure 4] FIG. 10 illustrates an example of speech recognition for an administrator's speech according to the present disclosure. [Figure 5]FIG. 10 illustrates an example of a discrepancy between an administrator and a respondent's utterances according to the present disclosure. [Figure 6] 10A and 10B are diagrams illustrating examples of inconsistencies between utterances and detection results according to the present disclosure. [Figure 7] FIG. 10 illustrates an example of a detection means according to the present disclosure. [Figure 8] FIG. 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION
[0010] (Description of a speech recognition device according to an embodiment) Fig. 1 is a block diagram showing the configuration of a speech recognition device according to the present disclosure. The speech recognition device 100 will be described with reference to Fig. 1. The speech recognition device 100 determines whether the content of a recognized utterance regarding entry into an access-restricted area such as a runway, port, or road is inconsistent with any of the utterances of the administrator, the utterances of the responder, or the behavior of the entering aircraft. The speech recognition device 100 is a device that outputs a contradiction if a contradiction is found.
[0011] A restricted access area is an area, such as a runway, where entry by other aircraft is restricted due to one-by-one entry. For example, a restricted access area may be a seaway at a harbor or a roadway. A manager is a person who manages multiple aircraft and monitors the restricted access area as a whole. A responder is a person who responds to instructions from the manager. A responder is a person operating an aircraft, such as a pilot. An aircraft may be an airplane, a car, a ship, or an aerial drone.
[0012] As shown in Fig. 1, the speech recognition device 100 includes an utterance acquisition unit 101, a recognition unit 102, a determination unit 103, and an output unit 104. The speech recognition device 100 is configured as an information processing device 800. As shown in Fig. 8, the information processing device 800 includes a processor 801 that executes a program to perform processing, and a memory 802 that stores the program. The information processing device 800 may be configured as a single device or as multiple devices. The information processing device 800 may be a cloud server that distributes and processes some or all of its functions.
[0013] The speech acquisition unit 101 acquires the speech of the administrator. The speech acquisition unit 101 also acquires the speech of the respondent. The speech of the administrator is acquired, for example, by acquiring the speech of a microphone that picks up the voice of the administrator. The speech of the respondent is acquired, for example, by acquiring the speech obtained by subtracting the speech of the administrator from the conversation between the administrator and the respondent. By acquiring the speech of either the administrator or the respondent, or both, and calculating the difference, the speech of each of the administrator and the respondent can be acquired.
[0014] The recognition unit 102 recognizes the content of the acquired utterance related to entry into the restricted access area. The recognition unit 102 can use artificial intelligence (AI) capable of speech recognition. The content related to entry is, for example, an administrator granting permission to enter the restricted access area. The content related to entry is, for example, a responder confirming that they will enter the restricted access area.
[0015] The determination unit 103 determines whether or not the recognized utterance contains a contradiction in content related to entry into the entry-restricted area. The contradiction in content related to entry may be determined by determining whether or not the administrator's utterance is contradictory to the administrator's next utterance. The contradiction in content related to entry may be determined by determining whether or not the content of the administrator's utterance is contradictory to the content of the respondent's utterance.
[0016] The speech recognition device 100 may include a detection unit. The detection unit detects the entry of an aircraft into an access-restricted area. The detection unit may be, for example, a sensor or a surveillance camera. The inconsistency with the content related to the entry here may be determined based on whether or not the administrator's speech is inconsistent with the detection result of the detection unit. The detection result of the detection unit may be the actual operation of the entering aircraft.
[0017] When the determination unit 103 determines that there is a contradiction, the output unit 104 outputs that there is a contradiction. The output unit 104 can output by means of audio output, display output on a display unit, generation of an alarm sound, or flashing a light, etc. The output unit 104 outputs to a display unit or alarm sound generator that can be recognized by the administrator, a supervisor who supervises the administrator, or a responder, or sends a warning to an external display device or sends a warning to an external warning system. The supervisor is the person who supervises the administrator and double-checks entry into the restricted access area.
[0018] The speech recognition device 100 may include a notification unit regarding entry into a restricted access area. The notification unit is, for example, a runway status light (RWSL). The notification unit recognizes the administrator's speech and issues a notification regarding entry into a restricted access area by changing the traffic light of an accessible runway to a green or other clearing indication. This allows the pilot to confirm the correct path and prevents accidents.
[0019] The utterance acquisition unit 101, the recognition unit 102, the determination unit 103, the output unit 104, the detection unit, and the presentation unit may be read as utterance acquisition means, recognition means, determination means, output means, detection means, and presentation means.
[0020] With this configuration, an utterance recognition device is provided that determines whether or not there is a contradiction in the content of a recognized utterance regarding entry into an entry-restricted area, and outputs a message indicating that there is a contradiction.
[0021] (Description of speech recognition method according to embodiment) 2 is a flowchart of a speech recognition method according to the present disclosure. Referring to FIG.
[0022] As shown in FIG. 2, first, an utterance is acquired (step S201). The utterance recognition device 100 acquires an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator. Next, the content is recognized (step S202). The utterance recognition device 100 recognizes the content of the acquired utterance related to entry into the access-restricted area. Next, it determines whether or not the content is contradictory (step S203). The utterance recognition device 100 determines whether or not the content of the recognized utterance related to entry into the access-restricted area is contradictory. If there is a contradiction (Yes in step S203), it outputs a fact that there is a contradiction (step S204) and ends the processing. The utterance recognition device 100 outputs a fact that there is a contradiction. If there is no contradiction (No in step S203), it ends the processing.
[0023] With this configuration, a speech recognition method is provided that determines whether or not the content of a recognized utterance regarding entry into an entry-restricted area is inconsistent, and outputs a message indicating that an inconsistency exists if there is one.
[0024] (Explanation of a specific example of speech recognition according to an embodiment) FIG. 3 is a diagram illustrating an exemplary outline of speech recognition according to the present disclosure. FIG. 4 is a diagram illustrating an example of speech recognition of an administrator's speech according to the present disclosure. FIG. 5 is a diagram illustrating an example of a contradiction between an administrator's speech and a respondent's speech according to the present disclosure. FIG. 6 is a diagram illustrating an example of a contradiction between an utterance and a detection result according to the present disclosure. FIG. 7 is a diagram illustrating an example of a detection means according to the present disclosure. Specific examples of speech recognition will be described with reference to FIGS. 3 to 7.
[0025] Figure 3 illustrates speech recognition in air traffic control. As shown in Figure 3, a controller manages a restricted access area. The controller's voice is transmitted to a console via a microphone, and then transmitted to a radio transmitter and receiver carried by a pilot. The pilot's voice is transmitted from the radio transmitter and receiver to the console, and then from the console to headphones or the like carried by the controller.
[0026] Location 1 for acquiring speech is between the console and the radio transmitter. This location transmits only the voice of the administrator, so the administrator's speech can be acquired. Location 2 for acquiring speech is the controller's microphone and headphones. This location picks up only the voice of the administrator, so the administrator's speech can be acquired.
[0027] Location 3 for acquiring speech is a radio receiver. This location can acquire speech from the pilot who responds and speech from the controller who manages the call. Alternatively, location 4 for acquiring speech is between the radio receiver and a DREC (Digital Recorder). This location can acquire speech from the pilot who responds and speech from the controller who manages the call.
[0028] The respondent's utterance can be obtained by subtracting the utterances at locations 1 and 2 from the utterances at locations 3 and 4 where the utterances are obtained.
[0029] As shown in Figure 4, the voice of the controller is transmitted between the console and the radio transmitter, and the voice of the controller and pilot is transmitted between the radio receiver and the console. The voices of the controller and pilot are then acquired and recognized using artificial intelligence. The text of the recognition results can be displayed on the display. For example, if the controller says, "RWY 16R at A14 Line up and wait XXX237" and the pilot says "XXX237 tokyo tower RWY16R at A14 Line up and wait " he responded by repeating.
[0030] Only the instructions for the aircraft to approach the runway are extracted from the textual content, and when an approach instruction is issued, the traffic light is changed from red to green. Since the above instruction is an instruction to approach the runway, the traffic light for taxiway A14 is changed from red to green to guide the aircraft. The traffic light is changed back to red after a certain period of time has passed.
[0031] The bottom diagram in Figure 4 shows what is displayed on the display. Here, it is advisable to highlight the instructions with an underline or the flight name in bold to make them easier to distinguish. Other highlighting may also be set to include the names of two way stations adjacent to the runway. The display identifies and displays the controller and pilot. The pilot confirms the voice instructions and signals and directs the aircraft into the restricted approach area.
[0032] FIG. 5 shows a conversation 500 between a controller and a pilot. Here, the controller is speaking with the call sign "RWY 16R at A14 Line up and wait XXX237". The pilot then read back, "XXX237 Tokyo Tower RWY16R at A14 hold short of The command was answered, "RWY16R," and the response was different from the instruction and the read-back.
[0033] In such a case, since there is a contradiction between the administrator's speech and the respondent's speech, the determination unit 103 can determine that the recognized speech has a contradiction in content related to entering the restricted area. In such a case, the output unit 104 issues an alarm or the like to call attention.
[0034] Similarly, if an utterance by an administrator is compared with a subsequent utterance by the administrator to cause multiple aircraft to enter a restricted access area, a contradiction occurs. In this case, the determination unit 103 can determine that the recognized utterances regarding the entry into the restricted access area are contradictory. In such a case, the output unit 104 also issues an alarm, for example.
[0035] As shown in Figure 6, a gate sensor that manages approach and exit to runway 603 may be installed on the taxiway adjacent to the runway to detect aircraft approach. Dialogue 601 instructs Xxx719 to approach A14. Even if gate sensor #1 detects 0, this is correct and no alarm is issued.
[0036] The conversation 602 is an instruction to another aircraft XXX237 to wait at A14. However, when detected by gate sensors #1→#0, it indicates that two aircraft are approaching the same runway 603. Therefore, the output unit 104 issues an alarm.
[0037] Also, suppose that after the conversation 601 has taken place, a vehicle 604 enters the runway 603. This can be detected by the gate sensor A12. In this case, the output unit 104 issues an alarm.
[0038] As shown in Figure 7, entry and exit to runway 702 may be detected by gate sensor 703 and surveillance camera 701. For example, when approaching the runway, an air traffic controller issues a command to "line up and wait." In this case, no permission is given to multiple aircraft to approach runway 702. Instead of "wait," a command to "clear to take off" may also be issued.
[0039] The instruction "cross RWY" may also be given. For example, "Xxx464 at L15 cross RWY16R taxi via K" is an instruction to cross the runway from L15 to K. In this case, two approach instructions cannot be given for one runway, so it is necessary to confirm that no further instructions are given until the aircraft receiving the instruction is thought to have exited the runway, in order to prevent other aircraft from making a mistaken approach. With such an instruction, the approach taxiway is specified, so the exit taxiway is also predicted. The information from this sensor is monitored.
[0040] If voice recognition does not issue instructions to runway 702 and gate sensors 703 from D-1 to D-7 detect an approach, it is an erroneous approach and an immediate warning is issued. The location can be identified by the number of the gate sensor 703. The location detected by the surveillance camera 701 linked to the gate sensor 703 can be instantly enlarged and displayed, eliminating the need to search with binoculars. The surveillance camera is, for example, a PTZ (pan, tilt, zoom) camera. Furthermore, if a confirmation landing aircraft is detected, an instruction to return can be immediately issued.
[0041] Some or all of the processes in the speech recognition device 100 described above can be implemented as a computer program. Such a program can be stored and provided to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible recording media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The program may also be provided to a computer by various types of temporary computer-readable media. Examples of temporary computer-readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable media can provide the program to a computer via a wired communication path such as an electric wire or optical fiber, or via a wireless communication path.
[0042] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.
[0043] Each drawing is merely an example for describing one or more embodiments. Each drawing may relate not only to one particular embodiment, but also to one or more other embodiments. As will be understood by those skilled in the art, various features or steps described with reference to any one drawing can be combined with features or steps shown in one or more other drawings to create, for example, an embodiment not explicitly shown or described. Not all features or steps shown in any one drawing are necessary to describe an exemplary embodiment, and some features or steps may be omitted. The order of steps described in any drawing may be changed as appropriate.
[0044] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes. (Appendix 1) a speech acquisition means for acquiring speech from a manager who manages an access-restricted area, which is an area with access restriction, or a respondent who responds to the manager; a recognition means for recognizing content of the acquired utterance related to entry into the access restriction area; a determination means for determining whether or not the recognized utterance has a contradiction in content related to entry into the entry-restricted area; an output means for outputting a signal indicating that there is a contradiction when the determination means determines that there is a contradiction. (Appendix 2) 2. The speech recognition device according to claim 1, wherein the determination means determines whether the content of the recognized utterance related to entry into the restricted access area is inconsistent depending on whether the utterance of the administrator is inconsistent with the next utterance of the administrator. (Appendix 3) The speech recognition device described in Appendix 1, wherein the determination means determines whether the content of the recognized utterance related to entry into the access-restricted area is contradictory by determining whether the content of the administrator's utterance and the content of the respondent's utterance are contradictory. (Appendix 4) a detection means for detecting entry into the access restriction area, 2. The speech recognition device of claim 1, wherein the determination means determines whether the content of the recognized utterance related to entry into the restricted access area is inconsistent by determining whether the administrator's utterance is inconsistent with the detection result of the detection means. (Appendix 5) 2. The speech recognition device of claim 1, wherein the output means outputs to a display or an alarm generating unit that can be recognized by the manager, a supervisor who supervises the manager, or the respondent, or sends a warning to an external display device, or sends a warning to an external warning system. (Appendix 6) The speech recognition device described in Appendix 5, wherein the output means highlights the content on the display unit when the administrator's speech or the respondent's speech contains content related to entering a restricted access area. (Appendix 7) The speech recognition device according to claim 1, further comprising a notification means for providing a notification regarding entry into the entry-restricted area. (Appendix 8) Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; A program that causes a speech recognition device to output a contradiction when it is determined that a contradiction exists. (Appendix 9) The program described in Appendix 8 determines whether the content of the recognized utterance related to entry into the restricted access area is contradictory based on whether the administrator's utterance is contradictory to the administrator's next utterance. (Appendix 10) The speech recognition device Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; The speech recognition method outputs a contradiction when it is determined that a contradiction exists.
[0045] Some or all of the elements (e.g., configurations and functions) described in Supplementary Note 2 to Supplementary Note 7 that are dependent on Supplementary Note 1 {e.g., device} may also be dependent on Supplementary Note 8 {e.g., program} and Supplementary Note 10 {e.g., method} in the same dependency relationship as Supplementary Note 2 to Supplementary Note 7. Some or all of the elements described in any Supplementary Note may be applied to various hardware, software, recording means for recording software, systems, and methods. [Explanation of symbols]
[0046] 100 speech recognition device, 101 speech acquisition unit, 102 recognition unit, 103 determination unit, 104 output unit
Claims
1. a speech acquisition means for acquiring speech from a manager who manages an access-restricted area, which is an area with access restriction, or a respondent who responds to the manager; a recognition means for recognizing content of the acquired utterance related to entry into the access restriction area; a determination means for determining whether or not the recognized utterance has a contradiction in content related to entry into the entry-restricted area; an output means for outputting a signal indicating that there is a contradiction when the determination means determines that there is a contradiction.
2. 2. The speech recognition device according to claim 1, wherein the determination means determines whether the recognized utterance has a contradiction in content related to entry into the restricted-entry area based on whether the utterance of the administrator is contradictory to a next utterance of the administrator.
3. 2. The speech recognition device according to claim 1, wherein the determination means determines whether or not the content of the recognized utterance related to entry into the restricted access area is contradictory by determining whether or not the content of the utterance of the administrator and the content of the utterance of the respondent are contradictory.
4. a detection means for detecting entry into the access restriction area, 2. The speech recognition device according to claim 1, wherein the determination means determines whether the content of the recognized utterance related to entry into the restricted access area is inconsistent by determining whether the administrator's utterance is inconsistent with the detection result of the detection means.
5. 2. The speech recognition device according to claim 1, wherein the output means outputs to a display unit or an alarm sound generating unit that can be recognized by the manager, a supervisor who supervises the manager, or the respondent, or transmits a warning to an external display device, or transmits a warning to an external warning system.
6. 6. The speech recognition device according to claim 5, wherein when the speech of the manager or the speech of the respondent includes content related to entering an access-restricted area, the output means highlights the content on the display unit.
7. The speech recognition device according to claim 1 , further comprising a notification unit that provides a notification regarding entry into the entry-restricted area.
8. Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; A program that causes a speech recognition device to output a contradiction when it is determined that a contradiction exists.
9. The program according to claim 8, wherein the program determines whether the content of the recognized utterance related to entry into the restricted access area is inconsistent based on whether the administrator's utterance is inconsistent with the administrator's next utterance.
10. The speech recognition device Acquire an utterance from an administrator who manages an access-restricted area, which is an area with access restrictions, or from a respondent who responds to the administrator; Recognizing content related to entry into the access restriction area in the acquired utterance; determining whether the recognized utterance has a contradiction in content related to entering the restricted access area; The speech recognition method outputs a contradiction when it is determined that a contradiction exists.
Citation Information
Patent Citations
Control operation support device and control operation instruction method
JP2012043200A