Information processing device, information processing method, and recording medium

Abstract

An information processing device (10) comprises: an acquisition means (50) that acquires series data; a likelihood ratio calculation means (60) that calculates a likelihood ratio matrix indicating the degree of matching between the series data with each of a plurality of registered data items; a determination means (70) that determines which of the plurality of registered data items matches the series data, on the basis of the likelihood ratio matrix; and an exclusion means (80) that determines whether the minimum value of each row in the likelihood ratio matrix is below a predetermined exclusion threshold, and that excludes the row falling under the exclusion threshold from the next calculation of the likelihood ratio matrix, and thereafter, in the likelihood ratio calculation means. According to such an information processing device, it is possible to suppress an increase of the processing load for likelihood ratio calculation.

Description

Information processing device, information processing method, and recording medium

[0001] The present disclosure relates to the technical fields of an information processing device, an information processing method, and a recording medium.

[0002] Known examples of this type of device include one that classifies sequential data using likelihood ratios. For example, Patent Document 1 discloses a device that sequentially acquires and analyzes multiple elements included in the sequential data to classify the sequential data into one of multiple predetermined classes.

[0003] International Publication No. 2020 / 194497

[0004] This disclosure aims to improve upon the related art discussed above.

[0005] One aspect of the information processing device disclosed herein includes an acquisition means for acquiring sequence data, a likelihood ratio calculation means for calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data, a determination means for determining which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, and an exclusion means for determining whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and for excluding rows below the exclusion threshold from the likelihood ratio calculation means' subsequent calculations of the likelihood ratio matrix.

[0006] One aspect of the information processing method disclosed herein involves using at least one computer to acquire sequence data, calculate a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data, determine which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, determine whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and exclude rows that are below the exclusion threshold from being included in subsequent calculations of the likelihood ratio matrix.

[0007] One aspect of the recording medium of this disclosure is a computer program recorded on at least one computer that causes the computer to execute an information processing method, which includes acquiring sequence data, calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data, determining which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, determining whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excluding rows that fall below the exclusion threshold from subsequent calculations of the likelihood ratio matrix.

[0008] 1 is a block diagram showing the hardware configuration of a first information processing device. FIG. 2 is a block diagram showing the functional configuration of the first information processing device. FIG. 3 is a diagram showing an example of a likelihood ratio matrix calculated by the first information processing device. FIG. 4 is a graph showing an example of a decision threshold used in the first information processing device. FIG. 5 is a graph showing an example of an exclusion threshold used in the first information processing device. FIG. 6 is a diagram showing an example of a class excluded by the first information processing device. FIG. 7 is a flowchart showing the operation flow of the first information processing device. FIG. 8 is a block diagram showing the functional configuration of a second information processing device. FIG. 9 is a flowchart showing the operation flow of the second information processing device. FIG. 10 is a conceptual diagram showing an example of an exclusion operation by a third information processing device. FIG. 11 is a conceptual diagram showing an example of an exclusion operation by a fourth information processing device. FIG. 12 is a side view showing an example of a gate system to which a fifth information processing device is applied.

[0009] Hereinafter, embodiments of an information processing device, an information processing method, and a recording medium will be described with reference to the drawings.

[0010] First Embodiment A first embodiment will be described with reference to FIGS. 1 to 7. FIG.

[0011] (Hardware Configuration) First, the hardware configuration of the first information processing apparatus will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the hardware configuration of the first information processing apparatus.

[0012] 1, a first information processing device 10 includes a processor 11, a RAM (Random Access Memory) 12, a ROM (Read Only Memory) 13, and a storage device 14. The information processing device 10 may further include an input device 15 and an output device 16. The processor 11, RAM 12, ROM 13, storage device 14, input device 15, and output device 16 are connected to each other via a data bus 17.

[0013] The processor 11 loads a computer program. For example, the processor 11 is configured to load a computer program stored in at least one of the RAM 12, the ROM 13, and the storage device 14. Alternatively, the processor 11 may load a computer program stored in a computer-readable storage medium using a storage medium reading device (not shown). The processor 11 may acquire (i.e., load) the computer program from a device (not shown) located outside the information processing device 10 via a network interface. The processor 11 controls the RAM 12, the storage device 14, the input device 15, and the output device 16 by executing the loaded computer program. In particular, in this embodiment, when the processor 11 executes the loaded computer program, a functional block that performs class classification based on likelihood ratios is realized within the processor 11. In other words, the processor 11 may function as a controller that executes each control in the information processing device 10.

[0014] The processor 11 may be configured as, for example, a central processing unit (CPU), a graphics processing unit (GPU), a field-programmable gate array (FPGA), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or a quantum processor. The processor 11 may be configured as one of these, or may be configured to use multiple processors in parallel.

[0015] The RAM 12 temporarily stores computer programs executed by the processor 11. The RAM 12 temporarily stores data that the processor 11 temporarily uses while it is executing the computer programs. The RAM 12 may be, for example, a dynamic random access memory (D-RAM) or a static random access memory (SRAM). Alternatively, other types of volatile memory may be used instead of the RAM 12.

[0016] The ROM 13 stores computer programs executed by the processor 11. The ROM 13 may also store fixed data. The ROM 13 may be, for example, a programmable read-only memory (PROM) or an erasable read-only memory (EPROM). Alternatively, other types of non-volatile memory may be used instead of the ROM 13.

[0017] The storage device 14 stores data that is to be saved long-term by the information processing device 10. The storage device 14 may operate as a temporary storage device for the processor 11. The storage device 14 may include, for example, at least one of a hard disk device, a magneto-optical disk device, an SSD (Solid State Drive), and a disk array device.

[0018] The input device 15 is a device that receives input instructions from a user of the information processing device 10. The input device 15 may include, for example, at least one of a keyboard, a mouse, and a touch panel. The input device 15 may be configured as a mobile terminal such as a smartphone or a tablet. The input device 15 may also be, for example, a device that includes a microphone and is capable of voice input.

[0019] The output device 16 is a device that outputs information related to the information processing device 10 to the outside. For example, the output device 16 may be a display device (e.g., a display) that can display information related to the information processing device 10. The output device 16 may also be a speaker or the like that can output information related to the information processing device 10 as audio. The output device 16 may be configured as a mobile terminal such as a smartphone or a tablet.

[0020] 1 shows an example of an information processing device 10 including multiple devices, but all or some of the functions may be realized by a single device. Such an information processing device may be configured to include only the above-mentioned processor 11, RAM 12, and ROM 13, and the other components (i.e., the storage device 14, input device 15, output device 16, etc.) may be provided by an external device connected to the information processing device 10. Furthermore, some of the calculation functions of the information processing device 10 may be realized by an external device (e.g., an external server, a cloud, etc.).

[0021] (Functional Configuration) Next, the functional configuration of the first information processing device 10 will be described with reference to Fig. 2. Fig. 2 is a block diagram showing the functional configuration of the first information processing device.

[0022] 2, the first information processing device 10 is a device that performs class classification of input sequence data, and is configured to include, as components for realizing its functions, an acquisition unit 50, a likelihood ratio calculation unit 60, a determination unit 70, and an exclusion unit 80. Each of the acquisition unit 50, the likelihood ratio calculation unit 60, the determination unit 70, and the exclusion unit 80 may be a processing block realized by, for example, the above-mentioned processor 11 (see FIG. 1).

[0023] The acquisition unit 50 is configured to be able to acquire sequence data. Sequence data here refers to data including multiple elements arranged in a predetermined order, such as time-series data. More specific examples of sequence data include, but are not limited to, video data, audio data, or subdivided image data. The acquisition unit 50 may acquire data directly from any data acquisition device (e.g., a camera, a microphone, etc.), or may read data that has been acquired in advance by a data acquisition device and stored in storage, etc. When acquiring data from a camera, the acquisition unit 50 may be configured to acquire data from each of multiple cameras.

[0024] The sequence data acquired by the acquisition unit 50 is input to the likelihood ratio calculation unit 60 together with a plurality of registered data. The registered data here is data registered in advance and is used for matching with the sequence data. The registered data is stored in, for example, a storage device. The storage device for storing the registered data may be provided by the information processing device 10 itself or an external device.

[0025] The likelihood ratio calculation unit 60 is configured to calculate a likelihood ratio based on the sequence data acquired by the acquisition unit 50 and multiple registered data. The "likelihood ratio" here is an index indicating the likelihood of a class to which the sequence data belongs. In this embodiment, the "likelihood ratio" is calculated by the acquisition unit 50 as an index indicating the degree of match between the sequence data and each of the multiple registered data. The likelihood ratio calculation unit 60 may be configured to calculate a likelihood ratio based on the relationship between at least two consecutive elements among multiple elements included in the sequence data. Hereinafter, the likelihood ratio (specifically, the likelihood ratio matrix) calculated by the likelihood ratio calculation unit 60 will be described with reference to FIG. 3 . FIG. 3 is a diagram illustrating an example of a likelihood ratio matrix calculated by the first information processing device.

[0026] Consider a likelihood ratio matrix in which likelihood ratios are arranged in a matrix, as shown in Figure 3. For ease of explanation, assume that the number of classification candidate classes (in other words, the number of registered data) is three: "Class 0," "Class 1," and "Class 2." p(X|y=0) is the likelihood that the sequential data is "Class 0." p(X|y=1) is the likelihood that the sequential data is "Class 1." p(X|y=2) is the likelihood that the sequential data is "Class 2."

[0027] In the first row from the top of the likelihood ratio matrix, the numerators of the log-likelihood ratios (hereinafter simply referred to as "likelihood ratios") are all p(X|y=0). In the second row from the top of the likelihood ratio matrix, the numerators of the likelihood ratios are all p(X|y=1). In the third row from the top of the likelihood ratio matrix, the numerators of the likelihood ratios are all p(X|y=2). On the other hand, in the first column from the left of the likelihood ratio matrix, the denominators of the likelihood ratios are all p(X|y=0). In the second column from the left of the likelihood ratio matrix, the denominators of the likelihood ratios are all p(X|y=1). In the third column from the left of the likelihood ratio matrix, the denominators of the likelihood ratios are all p(X|y=2).

[0028] The likelihood ratios on the diagonal of the likelihood ratio matrix (likelihood ratios shaded in gray in FIG. 3 ) have the same denominator and numerator. Specifically, the log{p(X|y=0) / p(X|y=0)} in the first row from the top and first column from the left, the log{p(X|y=1) / p(X|y=1)} in the second row from the top and second column from the left, and the log{p(X|y=2) / p(X|y=2)} in the third row from the top and third column from the left each have the same denominator and numerator. Furthermore, the likelihood ratios on opposite sides of the diagonal likelihood ratios have reversed denominators and numerators. Specifically, the numerator and denominator are reversed between log{p(X|y=0) / p(X|y=1)} in the first row from the top, second column from the left and log{p(X|y=1) / p(X|y=0)} in the second row from the top, first column from the left. Similarly, the numerator and denominator are reversed between log{p(X|y=0) / p(X|y=2)} in the first row from the top, third column from the left and log{p(X|y=2) / p(X|y=0)} in the third row from the top, first column from the left. The numerator and denominator are reversed between log{p(X|y=1) / p(X|y=2)} in the second row from the top, third column from the left and log{p(X|y=2) / p(X|y=1)} in the third row from the top, second column from the left. Therefore, likelihood ratios at opposite positions across the diagonal line have values ​​with opposite signs. In this way, the likelihood ratios in the likelihood ratio matrix are arranged like a skew matrix (antisymmetric matrix).

[0029] Returning to FIG. 1 , the determination unit 70 is configured to be able to classify sequential data based on the likelihood ratio matrix calculated by the likelihood ratio calculation unit 60. The determination unit 70 selects at least one class to which the sequential data belongs from among a plurality of classes that are classification candidates. Through this operation, the determination unit 70 determines which of a plurality of registered data the sequential data matches. The determination method used by the determination unit 70 will be specifically described below with reference to FIG. 4 . FIG. 4 is a graph showing an example of a determination threshold used in the first information processing device.

[0030] As shown in FIG. 4 , the determination unit 70 determines whether the minimum value of the likelihood ratios for each row in the likelihood ratio matrix exceeds a determination threshold. If the minimum value exceeds the determination threshold, the determination unit 70 determines that the class corresponding to that row (i.e., the registered data) matches the sequence data. In the example shown in FIG. 4 , the likelihood ratio of log{p(X|y=0) / p(X|y=1)} gradually increases and finally exceeds the determination threshold. In such a case, class 0 is determined to be the correct class (i.e., the registered data matches the sequence data).

[0031] Returning to FIG. 1 , the exclusion unit 80 is configured to determine whether the minimum value of each row in the likelihood ratio matrix calculated by the likelihood ratio calculation unit 60 is below a predetermined exclusion threshold. The exclusion unit 80 then excludes rows whose minimum value is below the exclusion threshold from the likelihood ratio calculation unit 60's subsequent calculations of the likelihood ratio matrix. The exclusion method used by the exclusion unit 80 will be described in detail below with reference to FIGS. 5 and 6 . FIG. 5 is a graph showing an example of an exclusion threshold used in the first information processing device. FIG. 6 is a diagram showing an example of classes excluded by the first information processing device.

[0032] As shown in FIG. 5 , the exclusion unit 80 determines whether the minimum value of the likelihood ratios for each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold. While the aforementioned determination threshold is a positive threshold (see FIG. 4 ), the exclusion threshold is a negative threshold. The exclusion threshold may be set to have the same absolute value as the determination threshold, or may be set to have a different absolute value. If there is a row whose minimum value is below the exclusion threshold, the exclusion unit 80 excludes the class (i.e., registered data) corresponding to that row from subsequent calculations of the likelihood ratio matrix. In the example shown in FIG. 5 , the likelihood ratio of log{p(X|y=1) / p(X|y=2)} gradually decreases and finally exceeds the determination threshold. In such a case, class 1 is excluded from subsequent calculations of the likelihood matrix.

[0033] As shown in Figure 6, when class 1 is excluded, all likelihood ratios of rows corresponding to class 1 are excluded from calculation from the next time onwards. Furthermore, because the likelihood ratio matrix is ​​an antisymmetric matrix, when rows corresponding to class 1 are excluded from calculation, columns corresponding to class 1 are also excluded from calculation.

[0034] (Flow of Operation) Next, the flow of operation of the first information processing device 10 will be described with reference to Fig. 7. Fig. 7 is a flowchart showing the flow of operation of the first information processing device.

[0035] 3, when the operation of the first information processing device 10 starts, the acquisition unit 50 first acquires sequence data (step S101). More specifically, the acquisition unit 50 sequentially acquires multiple elements included in the acquired sequence data.

[0036] Next, the sequence data acquired by the acquisition unit 50 and the read-out registered data are input to the likelihood ratio calculation unit 101 (step S102).The likelihood ratio calculation unit 60 then calculates a likelihood ratio matrix based on the sequence data and the registered data (step S103).

[0037] Next, the determination unit 70 determines whether there is a row in the likelihood ratio matrix whose minimum value exceeds the determination threshold (step S104). If it is determined that there is a row whose minimum value exceeds the determination threshold (step S104: YES), the determination unit 70 outputs a determination result that the class corresponding to that row is the correct class (step S105).

[0038] On the other hand, if it is determined that there is no row whose minimum value exceeds the determination threshold (step S104: NO), the exclusion unit 80 determines whether there is a row in the likelihood ratio matrix whose minimum value is below the exclusion threshold (step S106). If it is determined that there is a row whose minimum value is below the exclusion threshold (step S106: YES), the exclusion unit 80 excludes the class corresponding to that row from the calculation target for the next calculation onwards (step S107). On the other hand, if it is determined that there is no row whose minimum value is below the exclusion threshold (step S106: NO), the processing of the above-mentioned step S107 is omitted. In other words, the processing of excluding calculation targets is not executed.

[0039] Regardless of whether the calculation target has been excluded or not, the process is repeated from step S101. Thus, the process of obtaining elements included in the sequence data and calculating the likelihood ratio matrix is ​​repeatedly executed until it is determined that there is a row whose minimum value exceeds the determination threshold.

[0040] (Technical Effects) Next, technical effects obtained by the first information processing device 10 will be described.

[0041] As described with reference to FIGS. 1 to 7 , in the first information processing device 10, if a likelihood ratio matrix contains a minimum value below an exclusion threshold, the corresponding class is excluded from subsequent likelihood ratio calculations. This prevents the processing load from increasing when calculating a likelihood ratio matrix. For example, if there are N registered data items, an N×N likelihood ratio matrix must be calculated, which increases computational costs exponentially and may become a bottleneck for the overall processing of the device. However, in this embodiment, as described above, classes that fall below the exclusion threshold are excluded from subsequent calculations, thereby reducing computational costs by the amount of the excluded classes. Furthermore, because only classes that are easy to classify are excluded (in other words, classes whose likelihood ratios do not change significantly are not excluded), excluding a class from the calculations has little impact on overall accuracy. Therefore, the first information processing device 10 can efficiently reduce computational costs while maintaining high classification accuracy.

[0042] Second Embodiment A second embodiment will be described with reference to Figures 8 and 9. Note that the second embodiment differs from the first embodiment described above only in some configurations and operations, and other parts may be the same as the first embodiment. Therefore, the following will describe in detail the parts that differ from the first embodiment, and will omit a description of other overlapping parts as appropriate.

[0043] (Functional Configuration) First, the functional configuration of the second information processing device 10 will be described with reference to Fig. 8. Fig. 8 is a block diagram showing the functional configuration of the second information processing device. Note that in Fig. 8, the same elements as those shown in Fig. 2 are denoted by the same reference numerals.

[0044] 8, the second information processing device 10 is configured to include, as components for realizing its functions, an acquisition unit 50, a likelihood ratio calculation unit 60, a determination unit 70, an exclusion unit 80, and a feature integration unit 90. That is, the second information processing device 10 further includes a feature integration unit 80 in addition to the configuration described in the first embodiment (see FIG. 2). The feature integration unit 80 may be a processing block realized by, for example, the above-mentioned processor 11 (see FIG. 1).

[0045] The feature integration unit 80 is configured to integrate features of the sequential data and features of the registered data to generate integrated features. For example, when N pieces of registered data are registered, the feature integration unit 80 integrates the features of the sequential data with each of the N pieces of registered data to generate N integrated features. Note that the method for integrating the features is not particularly limited. For example, the feature combination unit 80 may integrate the features of the sequential data and the registered data by combining them. Alternatively, the feature combination unit 80 may integrate the features by calculating the difference between the features of the sequential data and the registered data.

[0046] The feature integration unit 80 may have a function of extracting features from at least one of the sequence data and the registered data. In this case, the feature integration unit 80 first extracts features from the sequence data and the registered data, and then integrates the extracted features of the sequence data with the features of the registered data. Note that features of the registered data may be extracted in advance and stored in the form of features.

[0047] The likelihood ratio calculation unit 60 in the second information processing device 10 is configured to calculate a likelihood ratio matrix using the integrated feature amounts integrated by the feature amount integration unit 80 described above.

[0048] (Operation Flow) Next, the operation flow of the second information processing device 10 will be described with reference to Fig. 9. Fig. 9 is a flowchart showing the operation flow of the second information processing device. Note that in Fig. 9, the same processes as those shown in Fig. 7 are denoted by the same reference numerals.

[0049] As shown in FIG. 9, when the operation of the second information processing device 10 is started, the acquisition unit 50 first acquires sequence data (step S101).

[0050] Next, the feature integration unit 80 extracts features from the sequence data acquired by the acquisition unit 50 (step S201). Note that the feature integration unit 80 may extract features from the registered data in the same way as from the sequence data. Thereafter, the feature integration unit 80 integrates the features of the sequence data and the registered data (step S202).

[0051] Next, the likelihood ratio calculation unit 60 calculates a likelihood ratio matrix based on the integrated features integrated by the feature integration unit (step S203). Note that the subsequent processing is the same as that described in the first embodiment (see FIG. 7), and therefore detailed description thereof will be omitted here.

[0052] (Technical Effects) Next, technical effects obtained by the second information processing device 10 will be described.

[0053] 8 and 9, the second information processing apparatus 10 generates integrated features by integrating the features of the sequence data and the features of the registered data. In this way, the integrated features can be used to calculate a likelihood ratio matrix, making it possible to appropriately determine the degree of match between the sequence data and the registered data.

[0054] Third Embodiment A third embodiment will be described with reference to Fig. 10. Note that the third embodiment differs from the first and second embodiments described above only in some of the operations, and other parts may be similar to the first and second embodiments. Therefore, the following will describe in detail the parts that differ from the embodiments already described, and will omit a description of other overlapping parts as appropriate.

[0055] (Exclusion Operation) First, the exclusion operation performed by the third information processing device 10 (i.e., the operation of the exclusion unit 80) will be described with reference to Fig. 10. Fig. 10 is a conceptual diagram showing an example of the exclusion operation performed by the third information processing device.

[0056] 10 , the likelihood ratio calculation unit 60 in the third information processing device 10 includes a classifier 65 that classifies classes. The classifier 65 is configured to output an N-dimensional logit when a D-dimensional feature quantity is input. More specifically, the classifier 65 calculates and outputs an N-dimensional logit by multiplying the input D-dimensional feature quantity by weights W_{ij} (i = [1, D], j = [1, N]).

[0057] The exclusion unit 80 in the third information processing device 10 deletes elements of classes to be excluded from the weights W_{ij} used by the above-mentioned classifier 65. As a result, if, for example, K classes are excluded, the weights W_{ij} (j=[1,N]) become weights W_{ij} (j=[1,N-K]) after the exclusion.

[0058] (Technical Effects) Next, technical effects obtained by the third information processing apparatus 10 will be described.

[0059] 10, in the third information processing device 10, elements of the class to be excluded are deleted from the weight W_{ij} used by the classifier 65. In this way, it is possible to reduce the load in the calculation process from the next time onwards by the amount of the excluded class.

[0060] <Fourth Embodiment> A fourth embodiment will be described with reference to Fig. 11. Note that the fourth embodiment differs from the first to third embodiments described above only in some of the operations, and other parts may be similar to the first to third embodiments. Therefore, the following will describe in detail the parts that differ from the embodiments already described, and will omit a description of other overlapping parts as appropriate.

[0061] (Exclusion Operation) First, the exclusion operation performed by the fourth information processing apparatus 10 will be described with reference to Fig. 11. Fig. 11 is a conceptual diagram showing an example of the exclusion operation performed by the fourth information processing apparatus.

[0062] 11 , the likelihood ratio calculation unit 60 in the fourth information processing device 10 includes a classifier 65 that classifies classes, as in the third embodiment described above. The exclusion unit 80 in the fourth information processing device 10 performs subsequent calculations using only the logits excluding the classes to be excluded from the N-dimensional logits output by the classifier 65. For example, if K classes are excluded, then subsequent calculations will be performed using only the logits for N−K classes.

[0063] (Technical Effects) Next, technical effects obtained by the fourth information processing apparatus 10 will be described.

[0064] 11, in the fourth information processing apparatus 10, subsequent calculations are performed using logits that exclude classes that are not to be calculated. In this way, the load on subsequent calculations can be reduced by the amount of the excluded classes.

[0065] Fifth Embodiment The fifth embodiment will be described with reference to Fig. 12. The fifth embodiment shows a specific application example of the first to fourth embodiments described above, and its configuration and operation may be generally similar to those of the first to fourth embodiments. Therefore, the following will describe in detail the parts that are different from the first to fourth embodiments already described, and will omit a description of other overlapping parts as appropriate.

[0066] (Gate System) First, a gate system to which the fifth information processing device 10 is applied will be described with reference to Fig. 12. Fig. 12 is a side view showing an example of a gate system to which the fifth information processing device is applied.

[0067] As shown in Fig. 12, the fifth information processing device 10 is applied to a gate system installed at the entrances of various facilities such as offices, apartment buildings, airports, event venues, and theme parks. The gate system includes a gate device 200 and a camera 300 as main components. The gate device 200 performs face authentication on an approaching target 500 to determine whether the target can pass through the gate. The camera 300 captures a facial image of the target to be used for face authentication. The information processing device 10 may be configured as a part of the gate device 300, for example, or may be configured by another server or the like.

[0068] When the gate system detects the approach of a target 500, it captures a facial image of the target 500 and performs facial authentication. The information processing device 10 performs facial authentication using this facial image. Specifically, the information processing device 10 acquires the facial image of the target as sequential data. Then, the information processing device 10 determines whether the facial image of the target 500 matches pre-registered enrollment data (i.e., facial images of registered users). More specifically, the information processing device 10 calculates a likelihood ratio matrix based on feature amounts extracted from the facial image of the target 500 and feature amounts of the enrollment data. Then, based on the calculated likelihood ratio matrix, it determines whether the face of the target 500 matches any face in the enrollment data. At this time, if the minimum value of each row of the likelihood ratio matrix is ​​below an exclusion threshold, the information processing device 10 excludes that class from subsequent calculation targets.

[0069] (Technical Effects) Next, technical effects obtained by the fifth information processing apparatus 10 will be described.

[0070] 12, the fifth information processing device 10 is used for face authentication using a face image. In face authentication, for example, a condition is imposed that authentication must be completed before the target 500 passes through a gate or before the target 500 moves out of the field of view of the camera 300, so authentication must be performed quickly and with high accuracy. However, when the face authentication system becomes large-scale (for example, when N is on the order of tens of thousands to hundreds of thousands), the calculation time increases, leading to a delay in the authentication speed.

[0071] However, according to the fifth information processing device 10, as described above, some classes are excluded from subsequent calculations depending on the determination result, so the processing load can be reduced accordingly. Also, the excluded classes are narrowed down to only those that are easy to determine, so the impact on the overall authentication accuracy is small. Therefore, according to the fifth information processing device 10, it is possible to realize a face authentication system with high authentication speed and high accuracy.

[0072] Although the above-described embodiment has been described as an example of authentication using a face image, the fifth information processing device 10 can also be applied to a device that acquires, as sequential data, images including parts other than the face (for example, an iris image, a fingerprint image, a palm print image, etc.) Furthermore, the fifth information processing device 10 can also be applied to a device that acquires, as sequential data, for example, audio other than images.

[0073] The scope of each embodiment also includes a processing method in which a program that operates the configuration of each embodiment to realize the functions of the above-described embodiments is recorded on a recording medium, the program recorded on the recording medium is read as code, and the program is executed on a computer. In other words, a computer-readable recording medium is also included in the scope of each embodiment. Furthermore, each embodiment includes not only a recording medium on which the above-described program is recorded, but also the program itself.

[0074] Examples of recording media that can be used include floppy disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, magnetic tapes, non-volatile memory cards, and ROMs. Furthermore, the scope of each embodiment is not limited to programs that execute processes by themselves, but also includes programs that execute processes by operating on an OS in conjunction with other software or expansion board functions. Furthermore, the program itself may be stored on a server, and part or all of the program may be downloadable from the server to a user terminal. The program may be provided to the user in, for example, a SaaS (Software as a Service) format.

[0075] <Supplementary Notes> The above-described embodiment may be further described as in the following supplementary notes, but is not limited to the following.

[0076] (Supplementary Note 1) The information processing device described in Supplementary Note 1 is an information processing device including: an acquisition means for acquiring sequence data; a likelihood ratio calculation means for calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data; a determination means for determining which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix; and an exclusion means for determining whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and for rows below the exclusion threshold, excluding them from the likelihood ratio calculation means' subsequent calculations of the likelihood ratio matrix.

[0077] (Supplementary Note 2) The information processing device described in Supplementary Note 2 is the information processing device described in Supplementary Note 1, further comprising a feature integration means that integrates features extracted from the sequence data and each of the features extracted from the plurality of registered data to generate a plurality of integrated features, and the likelihood ratio calculation means calculates the likelihood ratio matrix using the plurality of integrated features.

[0078] (Supplementary Note 3) The information processing device described in Supplementary Note 3 is the information processing device described in Supplementary Note 1 or 2, wherein the exclusion means deletes elements corresponding to rows that are below the exclusion threshold from weights used when the likelihood ratio calculation means calculates the likelihood ratio matrix.

[0079] (Supplementary Note 4) The information processing device described in Supplementary Note 4 is the information processing device described in Supplementary Note 1 or 2, wherein the exclusion means calculates the likelihood ratio matrix from the next time onwards by using the likelihood ratio matrix calculated by the likelihood ratio calculation means, excluding logits corresponding to rows that are below the exclusion threshold.

[0080] (Appendix 5) The information processing device described in Appendix 5 is the information processing device described in any one of Appendices 1 to 4, wherein the series data are facial images of a target, the plurality of registered data are facial images of registered users, and the determination means determines which of the registered users the target's face matches.

[0081] (Supplementary Note 6) The information processing method described in Supplementary Note 6 is an information processing method that, by at least one computer, acquires sequence data, calculates a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data, determines which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, determines whether or not the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excludes rows that are below the exclusion threshold from being targets for subsequent calculation of the likelihood ratio matrix by the likelihood ratio calculation means.

[0082] (Supplementary Note 7) The recording medium described in Supplementary Note 7 is a recording medium having recorded thereon a computer program for causing at least one computer to execute an information processing method of acquiring sequence data, calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data, determining which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, determining whether the minimum value of each row in the likelihood ratio matrix is ​​below an exclusion threshold, and excluding rows that are below the predetermined exclusion threshold from targets for subsequent calculation of the likelihood ratio matrix by the likelihood ratio calculation means.

[0083] (Supplementary Note 8) The computer program described in Supplementary Note 8 is a computer program that causes at least one computer to execute an information processing method of acquiring sequence data, calculating a likelihood ratio matrix indicating a degree of match between the sequence data and each of a plurality of registered data, determining which of the plurality of registered data the sequence data matches based on the likelihood ratio matrix, determining whether or not a minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excluding rows below the exclusion threshold from targets for subsequent calculation of the likelihood ratio matrix by the likelihood ratio calculation means.

[0084] This disclosure may be modified as appropriate within the scope that does not contradict the gist or idea of ​​the invention that can be read from the claims and the entire specification, and information processing devices, information processing methods, and recording media that involve such modifications are also included in the technical idea of ​​this disclosure.

[0085] REFERENCE SIGNS LIST 10 Information processing device 11 Processor 50 Acquisition unit 60 Likelihood ratio calculation unit 65 Classifier 70 Determination unit 80 Exclusion unit 90 Feature integration unit 200 Gate device 300 Camera 500 Object

Claims

1. an acquisition means for acquiring sequence data; a likelihood ratio calculation means for calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of a plurality of registered data; a determination means for determining which of the plurality of registered data sets the sequence data matches based on the likelihood ratio matrix; an exclusion means for determining whether or not the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excluding rows where the minimum value is below the predetermined exclusion threshold from subsequent calculations of the likelihood ratio matrix by the likelihood ratio calculation means; An information processing device comprising:

2. further comprising a feature integration means for integrating the feature extracted from the sequence data and each of the feature extracted from the plurality of registered data to generate a plurality of integrated feature quantities; the likelihood ratio calculation means calculates the likelihood ratio matrix using the plurality of integrated feature amounts. The information processing device according to claim 1 .

3. the exclusion means deletes elements corresponding to rows whose weights are below the exclusion threshold from weights used when the likelihood ratio calculation means calculates the likelihood ratio matrix.

3. The information processing device according to claim 1 or 2.

4. the exclusion means calculates the likelihood ratio matrix for the next time and thereafter by using the likelihood ratio matrix calculated by the likelihood ratio calculation means, excluding logits corresponding to rows below the exclusion threshold.

3. The information processing device according to claim 1 or 2.

5. the series of data are facial images of a target, and the plurality of registered data are facial images of registered users; The determination means determines whether the target face matches the face of any of the registered users.

3. The information processing device according to claim 1 or 2.

6. by at least one computer, Get the sequence data, calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of the plurality of registered data; determining which of the plurality of registered data sets the sequence data matches based on the likelihood ratio matrix; determining whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excluding rows where the minimum value is below the exclusion threshold from subsequent calculations of the likelihood ratio matrix; Information processing methods.

7. At least one computer Get the sequence data, calculating a likelihood ratio matrix indicating the degree of match between the sequence data and each of the plurality of registered data; determining which of the plurality of registered data sets the sequence data matches based on the likelihood ratio matrix; determining whether the minimum value of each row in the likelihood ratio matrix is ​​below a predetermined exclusion threshold, and excluding rows where the minimum value is below the exclusion threshold from subsequent calculations of the likelihood ratio matrix; A computer program that executes an information processing method.