Output device and sound output method

Abstract

This output device (10) comprises: a signal acquisition unit (11) that acquires a signal obtained by measuring a user performing a task; a first fatigue estimation unit (12a) for estimating a first fatigue level, which is the degree of fatigue based on the acquired signal, by using a preconstructed first estimation model indicating the correlation between the signal obtained from the user and the level of fatigue the user experiences up to a future first period; a second fatigue estimation unit (12b) for estimating a second fatigue level, which is the degree of fatigue based on the acquired signal, by using a preconstructed second estimation indicating the correlation between the signal obtained from the user and the level of fatigue the user experiences up to a future second period longer than the first period; and an output unit (14) for outputting recommendation information about the execution of a task by using the estimated first fatigue level and second fatigue level.

Description

Output device and output method

[0001] The present disclosure relates to an output device and an output method.

[0002] There has been a need for estimating a person's fatigue level. For example, if a user working can know his or her own fatigue level, it would be useful as information for dealing with the fatigue level, such as taking a rest at an appropriate time. For example, Patent Literature 1 discloses an information processing device that acquires a first fatigue level indicating the subject's fatigue level and a second fatigue level indicating the subject's fatigue level based on an index different from the first fatigue level, and displays information about the subject's fatigue.

[0003] International Publication No. 2022 / 049727

[0004] However, there is a need for users to know how the accumulation of fatigue will change in the future and what kind of impacts will result. Therefore, the present disclosure provides an output device or the like for informing users of the impacts of the accumulation of fatigue.

[0005] In order to solve the above problem, an output device according to one aspect of the present disclosure includes a signal acquisition unit that acquires a signal obtained by measuring a user performing a task; a first fatigue estimation unit that estimates a first fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed first estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a first period in the future; a second fatigue estimation unit that estimates a second fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed second estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a second period in the future that is longer than the first period; and an output unit that outputs recommendation information for performing the task using the estimated first fatigue level and second fatigue level.

[0006] In addition, an output method according to one aspect of the present disclosure is an output method executed by a computer, and includes the steps of: acquiring a signal obtained by measuring a user performing a task; estimating a first fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed first estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a first period in the future; estimating a second fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed second estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a second period in the future that is longer than the first period; and outputting recommendation information for performing the task using the estimated first fatigue level and second fatigue level.

[0007] According to the present disclosure, it is possible to inform the user of the effects of accumulated fatigue.

[0008] Fig. 1 is a diagram for explaining the function of an output device according to an embodiment. Fig. 2 is a diagram for explaining the function of an output device according to an embodiment. Fig. 3 is a block diagram showing the functional configuration of an output system according to an embodiment. Fig. 4 is a flowchart showing an output method according to an embodiment.

[0009] 1 and 2 are diagrams for explaining the functions of an output device according to an embodiment. Fig. 1 shows the connection relationships of functions related to general fatigue estimation, and Fig. 2 shows the connection relationships of functions related to fatigue estimation in this embodiment.

[0010] As shown in FIG. 1 , conventionally, two indices, such as a face image obtained by face image measurement and a heart rate signal obtained by heart rate measurement, are converted into feature quantities (here, a face feature quantity and a heart rate feature quantity), and the fatigue accumulated in a user is estimated from the two feature quantities based on the two indices, and the fatigue state of the user is displayed.

[0011] On the other hand, as mentioned in the Background Art section, there is a need among users to know how the accumulation of fatigue will change in the future and what kind of impacts will result. In other words, there are a certain number of users who want to know the time-series changes in the accumulation of fatigue and the associated impacts. Time-series changes in the accumulation of fatigue include fatigue that accumulates in the short term and fatigue that accumulates over the long term. By taking into account these types of fatigue that accumulate differently over time, and estimating and outputting the impacts that will occur on the user, it is possible to satisfy the above user needs.

[0012] Therefore, in this disclosure, the accumulation of fatigue in a user is divided into two time axes: fatigue that accumulates in the short term and fatigue that accumulates over the long term, and by taking these into consideration comprehensively, it is possible to perform a detailed analysis based on the progression of accumulated fatigue.

[0013] Specifically, as shown in FIG. 2 , a temporary fatigue estimation model is used to estimate short-term accumulated fatigue based on two feature quantities, such as facial feature quantities and heart rate feature quantities, and a fatigue estimation model is used to estimate long-term accumulated fatigue based on two feature quantities, such as facial feature quantities and heart rate feature quantities. Then, the estimated short-term accumulated fatigue and the estimated long-term accumulated fatigue are analyzed to determine how they affect the user. Here, the user is, for example, a user who performs a task. The content of the task is not particularly limited, but examples include sports or training to improve physical ability, desk tasks such as learning and work, and drivers operating moving objects such as vehicles, aircraft, and ships.

[0014] In this embodiment, the influence of accumulated fatigue on the user is analyzed to determine how the accumulated fatigue affects the task the user is performing, and a recommendation for continuing the task is output. In other words, rather than simply estimating fatigue and outputting the degree of fatigue, the output device analyzes how the fatigue affects the task the user is performing, and outputs the resulting recommendation information for continuing the task. In this way, the output device in this embodiment makes it possible to notify the user of how the accumulated fatigue will change in the future and what impact this will have.

[0015] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. Note that the embodiments described below all show comprehensive or specific examples of the present disclosure. Therefore, the numerical values, components, the arrangement and connection of the components, steps, and the order of steps shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Therefore, among the components in the following embodiments, components that are not described in the independent claims of the present disclosure will be described as optional components.

[0016] Furthermore, each figure is a schematic diagram and is not necessarily an exact illustration. Therefore, the scales and the like do not necessarily match in each figure. In each figure, the same reference numerals are used to denote substantially the same components, and redundant explanations will be omitted or simplified.

[0017] (Embodiment) [Configuration of Fatigue Level Estimation System] First, the functional configuration of the output system in this embodiment will be described in detail with reference to Fig. 3. Fig. 3 is a block diagram showing the functional configuration of the output system according to this embodiment.

[0018] As shown in FIG. 3, the output system according to the present embodiment includes a sensor 20, an output device 10, and a presentation device 30.

[0019] The sensor 20 is a device that measures the user performing a task and outputs the obtained signal.

[0020] As described above, the sensor 20 is a camera for measuring a facial image and a heart rate monitor for measuring a heart rate, and outputs a facial image and a heart rate signal as signals. The sensor 20 may be either the camera or the heart rate monitor. Other examples of the sensor 20 include a cerebral blood flow monitor, an electroencephalograph, a plethysmograph, a thermometer, a respirometer, and a thermal imaging camera for capturing facial temperature distribution. These may be used in combination with at least one of the camera and the heart rate monitor, or may be used in place of at least one of the camera and the heart rate monitor.

[0021] The output device 10 is a device that generates and outputs output information including recommendation information by analyzing a signal obtained from a user performing a task.

[0022] The output device 10 includes a signal acquisition unit 11, a fatigue estimation unit 12, an output content determination unit 13, and an output unit 14. The output device 10 includes a processor and a memory, and performs the above-mentioned several functions by executing the program by the processor using the program and information stored in the memory.

[0023] The signal acquisition unit 11 is a functional unit that acquires a signal output from the sensor 20. Based on the acquired signal, the signal acquisition unit 11 outputs a signal for analysis to the fatigue estimation unit 12. For example, the signal acquisition unit 11 has a function of converting the measurement signal acquired from the sensor 20 into feature quantities. In this case, the signal acquisition unit 11 converts the facial image, which is the measurement signal, into feature quantities for analysis, such as the degree of eye opening and closing and the degree of eye corner downturn, the eye closing time and number of blinks per unit time, the height of the corners of the mouth, the degree of mouth opening and closing, the amount of head movement and tilt, and facial color. In addition, the signal acquisition unit 11 converts the heart rate signal, which is the measurement signal, into feature quantities for analysis, such as heart rate and heart rate variability (RRI, SDNN, CVRR, RMSSD, LF, HF, LF / HF, TP, ccvTP, etc.). In addition, the signal acquisition unit 11 may convert the measurement signals output from the sensor 20, such as cerebral blood flow, brain waves, pulse waves, body temperature, respiratory rate and depth, and thermal images capturing facial temperature distribution, into feature quantities, which are signals for analysis different from these.

[0024] Furthermore, in the above description, a configuration has been described in which a measurement signal is acquired from the sensor 20 and converted into a signal for analysis by the signal acquisition unit 11 before being output to the fatigue estimation unit 12. However, the sensor 20 may convert the measurement signal before outputting it. In this case, the signal acquisition unit 11 acquires the converted signal for analysis as a signal, and can output the acquired signal as is to the fatigue estimation unit 12. In this way, the signal acquired by the signal acquisition unit 11 may be the measurement signal measured by the sensor 20 itself, or may be an analysis signal converted from the measurement signal.

[0025] The fatigue estimation unit 12 uses the analysis signal to estimate a first fatigue level corresponding to the degree of fatigue accumulated in the short term and a second fatigue level corresponding to the degree of fatigue accumulated in the long term, and outputs the estimation results to the output content determination unit 13.

[0026] As shown in the figure, the fatigue estimation unit 12 includes a first fatigue estimation unit 12a, a second fatigue estimation unit 12b, and a model construction unit 12c. The first fatigue estimation unit 12a and the second fatigue estimation unit 12b estimate the fatigue level using an estimation model constructed in advance by the model construction unit 12c. More specifically, the first fatigue estimation unit 12a uses the first estimation model constructed in advance by the model construction unit 12c to estimate a first fatigue level, which is the degree of fatigue that the user will experience within a first future period, based on the analysis signal. In other words, the short term refers to the period from the present to the first future period. The first estimation model is an estimation model that indicates a correlation between the signal obtained from the user and the degree of fatigue that the user will experience within the first future period.

[0027] Similarly, the second fatigue estimation unit 12b uses a second estimation model previously constructed by the model construction unit 12c to estimate a second fatigue level, which is the degree of fatigue the user will experience within a second future period, based on the analysis signal. In other words, "long-term" refers to the period from the present to the second future period. The second period is longer than the first period, for example, several to several dozen times longer than the first period. If the first period is several hours, the second period corresponds to several days, and if the first period is several days, the second period corresponds to several weeks to several months. The second estimation model is an estimation model that indicates a correlation between the signal obtained from the user and the degree of fatigue the user will experience within the second future period.

[0028] The signal obtained from the user may be input in the same way (the same signal) to both the first fatigue estimation unit 12 a and the second fatigue estimation unit 12 b, and fatigue levels for different periods may be estimated and output by the first fatigue estimation unit 12 a and the second fatigue estimation unit 12 b. Alternatively, the signal obtained from the user may include a plurality of signals, and different portions of the signal may be input to each of the first fatigue estimation unit 12 a and the second fatigue estimation unit 12 b, and fatigue levels for different periods may be estimated and output by the first fatigue estimation unit 12 a and the second fatigue estimation unit 12 b.

[0029] The first estimation model and the second estimation model are constructed by learning the relationship between the signals acquired by the signal acquisition unit 11 and the correct value of the user's fatigue level, which are obtained by the model construction unit 12c through previous measurements of the user. The correct value of the user's fatigue level includes, for example, a first fatigue index and a second fatigue index acquired by the model construction unit 12c and having different time constants. The first fatigue index has a smaller time constant than the second fatigue index and corresponds to the correct value of the first fatigue level. The second fatigue index has a larger time constant than the first fatigue index and corresponds to the correct value of the second fatigue level. The first fatigue index may be, for example, one or more of the user's short-term subjective fatigue level, salivary cortisol level, salivary α-amylase level, and blood lactate level. The second fatigue index may be, for example, one or more of the user's long-term subjective fatigue level, hair or nail cortisol level, and salivary immunoglobulin A level, but is not limited thereto.

[0030] Although the first estimation model and the second estimation model are described as machine learning models, the first estimation model and the second estimation model are not limited to this. The first estimation model and the second estimation model may be realized in any manner as long as they can output the magnitude of the first fatigue level and the magnitude of the second fatigue level from the analysis signal, and may be realized by a rule-based method using a threshold value or a statistical method, etc. Furthermore, the machine learning model is not particularly limited in its model structure, such as regression, decision tree, or neural network.

[0031] It has been mentioned that the model construction unit 12c uses a first fatigue index and a second fatigue index having different time constants when constructing the first estimation model and the second estimation model. However, the model construction unit 12c may also perform model learning using a single fatigue index with a single time constant, and then tune the time constant of the fatigue index in this model to increase or decrease it, thereby constructing the first estimation model and the second estimation model.

[0032] The output content determination unit 13 is a functional unit that determines the content to be output by selecting predetermined recommendation information for each combination based on the combination of the magnitude of the first fatigue level and the magnitude of the second fatigue level. The output content determination unit 13 also generates and outputs map information that visualizes the fatigue experienced by the user. The map information is, for example, information for identifying which part of the user's body the fatigue experienced by the user is attributed to and displaying that part in a manner that differentiates it from other parts. Alternatively, the map information may be information indicating the fatigue state, such as a graph or table showing the correlation between the first fatigue index and the second fatigue index. Thus, the map information is information that is different from recommendation information and is used to inform the user of the state of fatigue.

[0033] For example, the output content determination unit 13 generates output information including recommendation information and map information and outputs it to the output unit 14. By generating and outputting additional map information in addition to the recommendation information, the user can more accurately grasp their own fatigue state. The map information may be output instead of the recommendation information. In other words, it is sufficient that at least one of the recommendation information and the map information is output as output information.

[0034] The output unit 14 is a functional unit that converts the output information generated by the output content determination unit 13 and outputs it to the presentation device 30 so that the information can be presented (e.g., displayed) on the presentation device 30. For example, the output unit 14 converts the content of the output information into an image and outputs it to the presentation device 30.

[0035] The presentation device 30 is a device for presenting information output from the output device 10. The presentation device 30 is a device that presents information included in the output information by displaying an image on a display or the like, for example. When an image is output from the output device 10, the image is displayed on the presentation device 30. Alternatively, the presentation device 30 may be a device that presents information included in the output information by playing back audio.

[0036] [Operation of Fatigue Level Estimation System] Next, the operation of the output system will be described with reference to FIG. 4. FIG. 4 is a flowchart showing an output method according to an embodiment. As shown in FIG. 4, the output system first constructs an estimation model (S101). As described above, the estimation model is constructed by training a machine learning model using signals obtained by measuring the user in the past and indices (first fatigue index and second fatigue index) corresponding to the correct value of the user's fatigue level at that time. In the output device 10, the constructed estimation model is stored, for example, in a storage unit (not shown).

[0037] After the above model construction is performed in advance, the output system begins estimating fatigue levels and outputting information. First, the sensor 20 measures the user performing a task to obtain a measurement signal. The signal acquisition unit 11 acquires the measurement signal (S102) and converts it into a signal for analysis. In the fatigue estimation unit 12, the first fatigue estimation unit 12a estimates a first fatigue level using the first estimation model (S103). The first fatigue level is the level of fatigue that will occur from the present until a first period has elapsed, i.e., the degree of fatigue that will accumulate over the short term. Furthermore, the second fatigue estimation unit 12b estimates a second fatigue level using the second estimation model (S104). The second fatigue level is the level of fatigue that will occur from the present until a second period has elapsed, i.e., the degree of fatigue that will accumulate over the long term.

[0038] In this embodiment, it is possible to estimate the fatigue level accumulated in two time regions, the first period and the second period, in this way. Then, the output content determination unit 13 determines the content to be output (at least one of the user state and advice included in the recommendation information) based on the first fatigue level and the second fatigue level (S105), and generates map information and outputs output information including the map information.

[0039] The output unit 14 converts the output information, including the recommendation information and map information, into a corresponding image and outputs it to the presentation device 30 (S106), thereby causing the information contained in the output information to be displayed (presented) as an image on the presentation device 30.

[0040] [Example 1] Hereinafter, an example using the output system described above will be described by giving a specific example.

[0041] In Example 1 described below, the user is training for a sport as a task. In this case, the fatigue level can be considered as the training load. This makes it possible to manage the training load based on the magnitude of the first fatigue level and the magnitude of the second fatigue level. For example, the sensor 20 measures the user training for a sport to obtain a signal.

[0042] Here, if a period of several days is set as the first period and a period of several weeks to several months is set as the second period, if the fatigue level in the first period is high (for example, compared to a threshold value) and the fatigue level in the second period is also high (for example, compared to a threshold value), it can be said that the long-term load and the most recent load are both high, and that there is a possibility of overtraining beyond appropriate exercise (user state). Therefore, the output content determination unit 13 selects content (advice) to warn whether the user is overtraining and includes it in the output information as recommendation information. The recommendation information may also include the user state described above.

[0043] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is high (e.g., compared to a threshold), it can be said that the long-term load is high but the recent load is low, and tapering is occurring (user state). Tapering is a load adjustment method that sets a period for reducing the load toward a schedule that requires maximizing the results of the sports training, such as a match or tournament. Therefore, the output content determination unit 13 selects content (advice) recommending continuing the recent training and includes it in the output information as recommendation information. The recommendation information may also include the user state described above.

[0044] Furthermore, if the fatigue level in the first period is high (e.g., compared to a threshold) and the fatigue level in the second period is low (e.g., compared to a threshold), it can be said that the recent load is high but the long-term load is low (user state). Therefore, the output content determination unit 13 selects content (advice) recommending continuing the recent training and includes it in the output information as recommendation information. The recommendation information may also include the above-mentioned user state.

[0045] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is low (e.g., compared to a threshold), it can be said that the current load and the long-term load are low (user state). Therefore, the output content determination unit 13 selects content (advice) recommending an increase in training (load) and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0046] In this way, it is possible to grasp the possibility of overtraining, the tapering state, and the training state, and output information that will allow necessary improvements to be made.

[0047] [Example 2] In Example 2 described below, a user is working as a task. In this case, it is assumed that the working environment has been improved. In this case, the fatigue level can be considered as the effect of the improvement of the working environment. Then, it is possible to evaluate the effect of the improvement of the working environment based on the magnitude of the first fatigue level and the magnitude of the second fatigue level. For example, the sensor 20 measures the working user to obtain a signal.

[0048] Here, if a period of several days is set as the first period and a period of several weeks to several months is set as the second period, if the fatigue level in the first period is high (for example, compared to a threshold value) and the fatigue level in the second period is high (for example, compared to a threshold value), it can be said that fatigue has accumulated and the user is in a state of fatigue (user state), even if only temporarily. Therefore, the output content determination unit 13 selects content (advice) that indicates a state in which improvement to the working environment is necessary and that current improvement measures are not effective, and includes this content in the output information as recommendation information. The above user state may also be included as recommendation information.

[0049] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold value) and the fatigue level in the second period is high (e.g., compared to a threshold value), it can be said that the user is in a state (user state) where fatigue is accumulated but is not temporary. Therefore, the output content determination unit 13 selects content (advice) that indicates a state in which the working environment needs to be improved and that fatigue has been reduced through improvement measures, and includes this content in the output information as recommendation information. The recommendation information may also include the above-mentioned user state.

[0050] Furthermore, if the fatigue level in the first period is high (e.g., compared to a threshold value) and the fatigue level in the second period is low (e.g., compared to a threshold value), it can be said that the user is temporarily fatigued but not accumulated fatigue (user state). Therefore, the output content determination unit 13 selects content (advice) that indicates a state in which improvement of the working environment is not necessary and current improvement measures are not effective, and includes this content in the output information as recommendation information. The above user state may be included as recommendation information.

[0051] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is low (e.g., compared to a threshold), it can be said that the user is in a state (user state) where there is neither temporary nor accumulated fatigue. Therefore, the output content determination unit 13 selects content (advice) that indicates a state where improvement of the working environment is not necessary and where fatigue has been reduced through improvement measures, and includes this content in the output information as recommendation information. The above user state may be included as recommendation information.

[0052] This makes it possible to determine whether improvements to the working environment are necessary and the effectiveness of such improvements.

[0053] [Example 3] In Example 3 described below, a user is studying or working as a task. In this case, the fatigue level can be considered as an indicator of a state in which the user is unable to study or work efficiently. Therefore, it becomes possible to determine whether the user can study or work efficiently based on the level of the first fatigue level and the level of the second fatigue level. For example, the sensor 20 measures the user studying or working to obtain a signal.

[0054] Here, if a period of several hours is set as the first period and a period of several days is set as the second period, and the fatigue level during the first period is high (e.g., compared to a threshold value) and the fatigue level during the second period is high (e.g., compared to a threshold value), it can be said that drowsiness and a decrease in concentration are occurring and may continue in the future, resulting in a state (user state) in which efficiency is decreasing and mistakes are more likely to occur. Therefore, the output content determination unit 13 selects content (advice) recommending sufficient sleep and includes it in the output information as recommendation information. The recommendation information may also include the user state described above.

[0055] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is high (e.g., compared to a threshold), the user is not drowsy or lacking in concentration, but is in a state (user state) that is likely to occur in the future and is not suited to tasks that require thinking. Therefore, the output content determination unit 13 selects content (advice) recommending learning content or work content that requires relatively little thinking and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0056] Furthermore, if the fatigue level in the first period is high (for example, compared to a threshold value) and the fatigue level in the second period is low (for example, compared to a threshold value), it can be said that the user is in a state (user state) where drowsiness or a decrease in concentration is temporarily occurring. Therefore, the output content determination unit 13 selects content (advice) recommending a temporary break and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0057] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is low (e.g., compared to a threshold), it can be said that drowsiness or a decrease in concentration is unlikely to occur temporarily or in the future, and that the state (user state) is suitable for learning or work. Therefore, the output content determination unit 13 selects content (advice) informing the user that learning or work is going well and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0058] In this way, future plans can be understood based on the state of sleepiness and concentration.

[0059] [Example 4] In Example 4 described below, a user operates a mobile object as a task. In this case, the fatigue level can be considered as an indicator of a state that may cause an abnormality in the operation of the mobile object. Therefore, it is possible to determine whether or not an abnormality will occur in the operation of the mobile object based on the level of the first fatigue level and the level of the second fatigue level. For example, the sensor 20 measures the user operating the mobile object to obtain a signal.

[0060] Here, if a period of several hours is set as the first period and a period of several days is set as the second period, if the fatigue level during the first period is high (e.g., compared to a threshold value) and the fatigue level during the second period is high (e.g., compared to a threshold value), it can be said that drowsiness and a decrease in concentration are occurring and may continue in the future, indicating a dangerous state (user state) with a high probability of accidents involving moving objects. Therefore, the output content determination unit 13 selects content (advice) recommending taking a break and reviewing the driving plan for the next few days and includes it in the output information as recommendation information. The recommendation information may also include the user state described above.

[0061] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold value) and the fatigue level in the second period is high (e.g., compared to a threshold value), it can be said that the user is not experiencing drowsiness or a decrease in concentration, but is in a state (user state) that is likely to occur in the future. Therefore, the output content determination unit 13 selects content (advice) recommending a review of the driving plan for the next few days and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0062] Furthermore, if the fatigue level in the first period is high (for example, compared to a threshold value) and the fatigue level in the second period is low (for example, compared to a threshold value), it can be said that the user is in a state (user state) where drowsiness or a decrease in concentration is temporarily occurring. Therefore, the output content determination unit 13 selects content (advice) recommending a temporary break and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0063] Furthermore, if the fatigue level in the first period is low (e.g., compared to a threshold) and the fatigue level in the second period is low (e.g., compared to a threshold), it can be said that drowsiness or a decrease in concentration is unlikely to occur temporarily or in the future, and that the state (user state) is suitable for operating a mobile object. Therefore, the output content determination unit 13 selects content (advice) informing that there are no problems with operating the mobile object and includes it in the output information as recommendation information. The recommendation information may include the above-mentioned user state.

[0064] In this way, future plans can be understood based on the state of sleepiness and concentration.

[0065] [Effects, etc.] As described above, the output device 10 according to the first aspect includes a signal acquisition unit 11 that acquires signals obtained by measuring a user performing a task; a first fatigue estimation unit 12a that estimates a first fatigue level, which is the level of fatigue, based on the acquired signals, using a pre-constructed first estimation model that indicates a correlation between the signals obtained from the user and the level of fatigue that will occur in the user up to a first period in the future; a second fatigue estimation unit 12b that estimates a second fatigue level, which is the level of fatigue based on the acquired signals, using a pre-constructed second estimation model that indicates a correlation between the signals obtained from the user and the level of fatigue that will occur in the user up to a second period in the future that is longer than the first period; and an output unit 14 that outputs recommendation information for performing the task, using the estimated first fatigue level and second fatigue level.

[0066] This output device 10 can output recommendation information for performing a task based on two fatigue levels, the first fatigue level and the second fatigue level, which have different fatigue durations. Because the first fatigue level and the second fatigue level have different fatigue durations, the timing of their impact on the performance of the task and the manner in which they are addressed are also different. In other words, recommendation information for performing a task based on these two temporally different fatigue levels is more likely to be appropriate than recommendation information based on a single fatigue level. Thus, this aspect makes it possible to inform the user of the impact of the accumulation of fatigue.

[0067] The output device 10 according to a second aspect is the output device 10 according to the first aspect, wherein the second period is several times to several tens of times longer than the first period.

[0068] This makes it possible to output recommendation information for performing a task based on two fatigue levels that occur over periods that differ by several to several dozen times.

[0069] The output device 10 according to the third aspect is the output device 10 according to the first or second aspect, and further includes a model construction unit 12c that constructs at least one of a first estimation model and a second estimation model based on signals obtained by measuring a user in the past.

[0070] This allows the first fatigue level and the second fatigue level to be estimated using at least one of the first estimation model and the second estimation model constructed in the model construction unit 12c.

[0071] Furthermore, the output device 10 according to the fourth aspect is the output device 10 according to the third aspect, in which the model construction unit 12c constructs a first estimation model and a second estimation model using a signal obtained by measuring the user in the past and a first fatigue index and a second fatigue index having different time constants.

[0072] This makes it possible to use the first estimation model and the second estimation model constructed from signals obtained by measuring the user in the past and the first fatigue index and the second fatigue index having different time constants.

[0073] Furthermore, the output device 10 according to the fifth aspect is the output device 10 according to the fourth aspect, wherein the first fatigue index includes at least one of subjective fatigue level, salivary cortisol level, salivary alpha-amylase level, and blood lactate level.

[0074] This allows the use of a first estimation model constructed using a first fatigue index including at least one of subjective fatigue level, salivary cortisol level, salivary alpha amylase level, and blood lactate level.

[0075] Furthermore, the output device 10 according to a sixth aspect is the output device 10 according to the fourth or fifth aspect, wherein the second fatigue index includes at least one of subjective fatigue level, hair or nail cortisol level, and saliva immunoglobulin A level.

[0076] This allows the use of a second estimation model constructed using a second fatigue index including at least one of subjective fatigue level, hair or nail cortisol level, and saliva immunoglobulin A level.

[0077] Furthermore, the output device 10 according to a seventh aspect is the output device 10 according to any one of the first to sixth aspects, in which the task is sports training performed by the user, and the recommendation information includes information regarding tapering of the training.

[0078] This makes it possible to output recommendation information including information regarding tapering of training for the sports training being performed by the user.

[0079] An output device 10 according to an eighth aspect is the output device 10 according to any one of the first to seventh aspects, wherein the signal includes at least one of a heartbeat signal of the user and a facial image of the user.

[0080] According to this, at least one of the user's heart rate signal and the user's face image can be acquired as a signal to estimate the first fatigue level and the second fatigue level.

[0081] Furthermore, the output device 10 according to a ninth aspect is the output device 10 according to any one of the first to eighth aspects, in which the output unit 14 further outputs map information that visualizes fatigue experienced by the user.

[0082] This makes it possible to further output map information that visualizes the fatigue experienced by the user.

[0083] In addition, an output method according to a tenth aspect of this embodiment is an output method executed by a computer, and includes the steps of: acquiring a signal obtained by measuring a user performing a task; estimating a first fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed first estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that the user will experience by a first period in the future; estimating a second fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed second estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that the user will experience by a second period in the future that is longer than the first period; and outputting recommendation information for performing the task using the estimated first fatigue level and second fatigue level.

[0084] This can provide the same effects as the output device 10 described above.

[0085] (Other Embodiments) While the output device and output method according to the present disclosure have been described above based on the above-mentioned embodiments, the present disclosure is not limited to the above-mentioned embodiments. For example, the present disclosure also includes forms obtained by applying various modifications to the embodiments that would occur to those skilled in the art, and forms realized by arbitrarily combining the components and functions of the embodiments within the scope of the present disclosure.

[0086] Furthermore, for example, the present disclosure can be realized not only as an output device, but also as a program including, as steps, processes performed by each component of the output device, and as a computer-readable recording medium on which the program is recorded. The program may be pre-recorded on the recording medium, or may be supplied to the recording medium via a wide area network including the Internet.

[0087] In other words, the above-described comprehensive or specific aspects may be realized as a system, an apparatus, an integrated circuit, a computer program, or a computer-readable recording medium, or may be realized as any combination of a system, an apparatus, an integrated circuit, a computer program, and a recording medium.

[0088] REFERENCE SIGNS LIST 10 Output device 11 Signal acquisition unit 12 Fatigue estimation unit 12a First fatigue estimation unit 12b Second fatigue estimation unit 12c Model construction unit 13 Output content determination unit 14 Output unit 20 Sensor 30 Presentation device

Claims

1. An output device comprising: a signal acquisition unit that acquires signals obtained by measuring a user performing a task; a first fatigue estimation unit that estimates a first fatigue level, which is the level of fatigue based on the acquired signals, using a pre-constructed first estimation model that indicates a correlation between the signals obtained from the user and the level of fatigue that will occur in the user by a first period in the future; a second fatigue estimation unit that estimates a second fatigue level, which is the level of fatigue based on the acquired signals, using a pre-constructed second estimation model that indicates a correlation between the signals obtained from the user and the level of fatigue that will occur in the user by a second period in the future that is longer than the first period; and an output unit that outputs recommendation information for performing the task using the estimated first fatigue level and second fatigue level.

2. The output device according to claim 1, wherein the second period is several to several dozen times longer than the first period.

3. The output device according to claim 1, further comprising a model construction unit that constructs at least one of the first estimation model and the second estimation model based on a signal obtained by measuring the user in the past.

4. An output device as described in claim 3, wherein the model construction unit constructs the first estimation model and the second estimation model using signals obtained by measuring the user in the past and a first fatigue index and a second fatigue index having different time constants.

5. An output device according to claim 4, wherein the first fatigue index includes at least one of subjective fatigue level, salivary cortisol level, salivary alpha amylase level, and blood lactate level.

6. The output device according to claim 4, wherein the second fatigue index includes at least one of subjective fatigue level, hair or nail cortisol level, and saliva immunoglobulin A level.

7. An output device according to any one of claims 1 to 6, wherein the task is sports training performed by the user, and the recommendation information includes information regarding tapering of the training.

8. An output device according to any one of claims 1 to 6, wherein the signal includes at least one of a heartbeat signal of the user and a facial image of the user.

9. The output device according to any one of claims 1 to 6, wherein the output section further outputs map information that visualizes fatigue experienced by the user.

10. An output method executed by a computer, comprising: a step of acquiring a signal obtained by measuring a user performing a task; a step of estimating a first fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed first estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a first period in the future; a step of estimating a second fatigue level, which is the level of fatigue based on the acquired signal, using a pre-constructed second estimation model that indicates a correlation between the signal obtained from the user and the level of fatigue that will occur in the user by a second period in the future that is longer than the first period; and a step of outputting recommendation information for performing the task using the estimated first fatigue level and second fatigue level.

Images