Information processing apparatus, information processing method, and program

Abstract

To assist a user in recognizing subjects that meet a plurality of search conditions.SOLUTION: An information processing device is provided, comprising confidence level determination means configured to determine a lower limit of a confidence level indicative of a probability that a portrait image meets an appearance attribute for each of multiple appearance attributes entered by a user as search conditions, and search means for searching for portrait images with confidence levels exceeding the lower limit determined for each of the multiple appearance attributes.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a recording medium, and more particularly to an information processing apparatus, an information processing method, and a recording medium that execute a search process based on search conditions.

Background Art

[0002] Patent Document 1 describes an information search method for searching for an image that matches search conditions using one or more input keywords as search conditions. In Patent Document 1, for each image obtained as a result of the search, one search condition (that is, one keyword) that the image satisfies is presented. Therefore, the user can easily know based on which search condition the image was obtained.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the information search method described in Patent Document 1, when a search using a plurality of search conditions is executed, the user cannot know the object that satisfies two or more search conditions that are all or a part of the plurality of search conditions.

[0005] An object of the present invention is to help know an object that satisfies a plurality of search conditions.

Means for Solving the Problems

[0006] An information processing apparatus according to an aspect of the present invention includes: confidence determination means for determining, for each of a plurality of appearance attributes input as search conditions by a user, a lower limit value of a confidence level indicating the probability that a person image satisfies the attribute; and search means for performing a search for a person image that exceeds the lower limit value determined for each of the plurality of appearance attributes.

[0007] An information processing method according to an aspect of the present invention determines, for each of a plurality of appearance attributes input as search conditions by a user, a lower limit value of a confidence level indicating the probability that a person image satisfies the attribute, and performs a search for a person image that exceeds the lower limit value determined for each of the plurality of appearance attributes.

[0008] A program according to an aspect of the present invention causes a computer to perform a process of determining, for each of a plurality of appearance attributes input as search conditions by a user, a lower limit value of a confidence level indicating the probability that a person image satisfies the attribute, and a process of performing a search for a person image that exceeds the lower limit value determined for each of the plurality of appearance attributes.

Advantages of the Invention

[0009] According to the present invention, it is possible to help know a target that satisfies a plurality of search conditions.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Mode for Carrying Out the Invention

[0011] 〔Embodiment 1〕 With reference to FIGS. 1 to 6, the configuration of the information processing apparatus 1 according to the present Embodiment 1 will be described.

[0012] (Information Processing Apparatus 1) FIG. 1 is a block diagram showing the configuration of the information processing apparatus 1 according to the present Embodiment 1. As shown in FIG. 1, the information processing apparatus 1 includes a search unit 10 and a presentation unit 20.

[0013] The search unit 10 searches for a target that satisfies at least one of the one or more input search conditions. The search unit 10 is an example of a search means.

[0014] Specifically, the search unit 10 uses one or more keywords, images, etc. input by the user as search conditions to search for targets that meet each search condition from a database (not shown). For example, when the search target is image data of a person, the search conditions may be keywords representing one or more attributes of the person (such as gender, age, hair length, hair color, type of clothing, belongings, etc.).

[0015] The search unit 10 presents a search screen and accepts input of search conditions from the user. For example, when the user uses an input device such as a keyboard and a mouse to input one or more keywords on the search screen, the search unit 10 acquires the one or more input keywords as one or more search conditions.

[0016] The search unit 10 uses the one or more acquired search conditions to acquire one or more image data that match each search condition from the database. In the first embodiment, the image data stored in the database is the target of the search by the search unit 10. The search result by the search unit 10 includes one or more image data that satisfy at least one of the one or more input search conditions. Note that the information processing apparatus 1 may include a database or may be connected to a database via a network.

[0017] The search unit 10 transmits the above-described search result to the presentation unit 20. The search result includes at least one or more image data that satisfy at least one of the one or more input search conditions and information for identifying each target. For example, when the search target is image data stored in the database, the search result includes an image ID (Identification) assigned to each image data in the database.

[0018] Based on the search result by the search unit 10, the presentation unit 20 presents in a list which search conditions the target satisfies and which search conditions it does not satisfy among the one or more search conditions. The presentation unit 20 is an example of a presentation means.

[0019] Specifically, the presentation unit 20 receives the search results from the search unit 10. Based on the search results received from the search unit 10, the presentation unit 20 generates a list indicating which search conditions are satisfied and which are not satisfied among one or more search conditions for the target. The presentation unit 20 presents the list thus generated by outputting it to a display (not shown) or the like. The presentation unit 20 may present the list by any method.

[0020] For example, the presentation unit 20 presents the list using a bar graph assigned to each search condition. Alternatively, the presentation unit 20 presents a two-dimensional graph on the display, where one axis corresponds to the search condition and the other axis corresponds to the target of the search. In this graph, which search conditions are satisfied and which are not satisfied among one or more search conditions for the target are indicated by numbers, symbols, characters, figures, or pictures. Some examples of the presentation by the presentation unit 20 will be described later.

[0021] (Operation flow of the information processing apparatus 1) Referring to FIG. 2, the operation flow of the information processing apparatus 1 according to Embodiment 1 will be described. FIG. 2 is a flowchart showing the flow of processing executed by each part of the information processing apparatus 1.

[0022] As shown in FIG. 2, the search unit 10 searches for a target that satisfies at least one of the input one or more search conditions (S1). The search unit 10 transmits the search results to the presentation unit 20. As described above, the search results include at least one or more image data that satisfy at least one of the input one or more search conditions and information for identifying each target.

[0023] The presentation unit 20 receives the search results from the search unit 10. Based on the search results received from the search unit 10, the presentation unit 20 presents in a list which search conditions each target satisfies and which are not satisfied (S2). For example, the presentation unit 20 presents the list using a bar graph assigned to each search condition.

[0024] With the above, the operation of the information processing apparatus 1 ends.

[0025] (Example of search screen) Referring to FIG. 3, the search unit 10 will explain a search screen for receiving an input of search conditions from the user. FIG. 3 shows an example of a search screen displayed on a display or the like. The search screen shown in FIG. 3 includes a search box. One or more keywords are input into the search box as one or more search conditions.

[0026] In FIG. 3, three keywords, "male, black hair, blue shirt", are shown in the search box. When the user presses or clicks the "Search" key to the right of the search box, the search unit 10 executes a search using the three keywords input into the search box.

[0027] The search unit 10 performs a so-called OR search. That is, the search unit 10 searches the database for image data of a person classified into the attributes indicated by any of the three keywords, "male", "black hair", and "blue shirt", shown in FIG. 3.

[0028] The search unit 10 may use supervised learning or unsupervised learning to search for image data of a person classified into the attributes indicated by the keyword.

[0029] Specifically, the search unit 10 first identifies the feature data corresponding to the input keyword. For example, the search unit 10 identifies the feature data corresponding to the input keyword by referring to a table (not shown) that associates the keyword with the feature data.

[0030] Alternatively, the search unit 10 collects sample images tagged with the input keyword from a server on an arbitrary network. Then, the search unit 10 extracts feature data from the collected sample images. The feature data includes, for example, a feature vector. The search unit 10 associates the feature data extracted from the sample images with the input keyword.

[0031] Next, the search unit 10 extracts feature data from the image data on the database. Alternatively, the search unit 10 may extract the feature data in advance from the image data on the database. The search unit 10 calculates the similarity between the feature data extracted from the image data on the database and the feature data corresponding to the keyword. The similarity may be based on, for example, the distance between the feature vectors included in the feature data.

[0032] The search unit 10 extracts the image data whose calculated similarity exceeds the threshold value, or the top several (two or three) pieces of image data in descending order of similarity, from all the image data on the calculation database. The search unit 10 outputs the image data thus extracted as the result of the search.

[0033] Alternatively, the image data of the persons in the database may be classified or clustered in advance for each attribute of the persons. For example, information indicating the attributes of the persons is associated with the image data of the persons in the database, respectively.

[0034] The search unit 10 identifies the attribute corresponding to the input keyword by referring to a table (not shown) in which the keyword and the attributes of the persons are associated. The information indicating the attributes of the persons indicated by the input keyword is compared with the information indicating the attributes of the persons associated with the image data of the persons. Then, the search unit 10 searches the database for the image data of the persons classified into the same attributes as the attributes of the persons indicated by the keyword.

[0035] The presentation unit 20 presents a list of the image data of the persons obtained as the result of the search. For example, the presentation unit 20 presents the list of the image data of the persons obtained as the result of the search on a display (not shown).

[0036] Figure 3 shows, as an example of the search results, the image data of six persons. The image data of the six persons satisfies any of the search conditions. Different numbers ((1) to (6)) are shown below each image data. These numbers represent the image IDs assigned to the image data of the respective persons.

[0037] Here, the "attributes" of a person refer to the external characteristics of the person. In Figure 3, the list of attributes includes, as examples, gender, hair color, clothing color, clothing type, necktie, and items carried. However, the attributes of a person are not limited to these. In Figure 3, the symbol "black (5)" regarding hair color, which is one of the attributes, indicates that there are 5 persons with black hair included in the search results. That is, "black" represents a characteristic regarding one attribute, and the number within the parentheses represents the number of persons.

[0038] As described above, the presentation unit 20 presents, in a list, which of the one or more search conditions the target satisfies and which search conditions it does not satisfy based on the search results. Some examples of the list presentation by the presentation unit 20 will be described below.

[0039] (Example 1 of list presentation) FIG. 4 shows an example of a list presented by the presentation unit 20. FIG. 4 is a chart including three bar graphs assigned for each search condition. The three bar graphs shown in FIG. 4 correspond to the three input search conditions (i.e., "male", "black hair", "blue shirt"). The numerical values (90%, 80%, and 60% respectively) shown in the parentheses beside each search condition represent the lower limit values of the confidence level. The confidence level represents the probability that the target satisfies the search condition. The larger the numerical value, the higher the probability that the target satisfies the search condition (i.e., the more likely). For example, all targets included in the search results with the attribute "male" as the search condition have a confidence level of 90% or more. In other words, a target with a confidence level lower than 90% for being classified as "male" will not appear on the search screen shown in FIG. 3. In the present Embodiment 1, the lower limit value of the confidence level is set to a predetermined value in advance. However, as described in Embodiment 2, the lower limit value of the confidence level may be variable.

[0040] In the bar graphs shown in FIG. 4, whether a plurality of search conditions are satisfied by the targets included in the search results is represented by the presence or absence of any pattern of diagonal lines, vertical lines, or horizontal lines in the bar graph.

[0041] For example, in the top bar graph shown in FIG. 4, only the portions corresponding to the three targets specified by the image IDs (1) to (3) on the horizontal axis are added with diagonal lines. From this, the three targets specified by the image IDs (1) to (3) are classified as the search condition "male". In other words, the top bar graph in the chart of FIG. 4 represents that the three image data specified by the image IDs (1) to (3) are male images with a probability (i.e., the lower limit value of the confidence level, that is, 90%) or more.

[0042] (Example of list presentation 2) Figure 5 shows another example of the list presented by the presentation unit 20. Figure 5 is a chart in which bar graphs assigned for each search condition are connected in the vertical direction of the drawing. Figure 5 shows whether one object corresponding to the bar graph satisfies each of the three input search conditions. Also in Figure 5, similar to Figure 4, the numerical values shown beside the three search conditions represent the lower limit values of the confidence level.

[0043] In Figure 5, the six bar graphs respectively correspond to the six searched objects (image IDs (1) to (6)). For example, the leftmost bar graph corresponds to the image data of the object specified by image ID (1).

[0044] In Figure 5, in the leftmost bar graph, the parts corresponding to the three search conditions (i.e., "male", "black hair", "blue shirt") are represented by double lines. From this, it can be seen that the object specified by image ID (1) satisfies all of the above three search conditions.

[0045] On the other hand, in Figure 5, in the second bar graph from the left, only the parts corresponding to the two search conditions (i.e., "male", "black hair") are represented by double lines, and the part corresponding to the remaining one search condition (i.e., "blue shirt") is represented by a single line. From this, it can be seen that the object specified by image ID (2) satisfies only the two search conditions (i.e., "male", "black hair").

[0046] (Example 3 of list presentation) Figure 6 shows yet another example of the list presented by the presentation unit 20. Figure 6 includes six charts assigned for each object. The six charts shown in Figure 6 respectively correspond to the six objects (image IDs (1) to (6)) shown in Figure 3. Also in Figure 6, similar to Figures 4 and 5, the numerical values beside the input search conditions represent the lower limit values of the confidence level.

[0047] In addition, the positions, shapes, and sizes of the six charts correspond to the positions, shapes, and sizes of the six targets in the search screen shown in FIG. 3. Therefore, the user can easily understand which person's image data each of the six charts shown in FIG. 6 corresponds to.

[0048] In FIG. 6, whether the target satisfies which search condition and which search condition it does not satisfy among one or more search conditions is represented by diagonal lines, vertical lines, or horizontal lines in the chart. For example, a vertical line is given to the lower right chart corresponding to the target specified by the image ID (6). This vertical line corresponds to the search condition of "black hair". From this, it can be seen that the target specified by the image ID (6) satisfies only the search condition of "black hair" with at least the certainty of the lower limit value of the certainty.

[0049] (Effect of this Embodiment) The search unit 10 searches for targets that satisfy one or more input search conditions, and the presentation unit 20 presents in a list which search conditions the target satisfies and which search conditions it does not satisfy among one or more search conditions based on the result of the search by the search unit 10. When a plurality of search conditions are input, the user can check the presented list and understand the targets that satisfy two or more search conditions that are all or part of the plurality of search conditions. Therefore, it can help to know the targets that satisfy a plurality of search conditions.

[0050] [Embodiment 2] Referring to FIGS. 7 to 9, Embodiment 2 will be described.

[0051] (Information Processing Apparatus 2) FIG. 7 is a block diagram showing the configuration of the information processing apparatus 2 according to Embodiment 2. As shown in FIG. 7, the information processing apparatus 2 includes a search unit 10, a presentation unit 20, and a certainty determination unit 30. That is, the information processing apparatus 2 according to Embodiment 2 is different in configuration from the information processing apparatus 1 according to Embodiment 1 in that it includes a certainty determination unit 30.

[0052] The confidence determination unit 30 determines the lower limit value of the confidence, which is a value representing the likelihood of the search result. The confidence determination unit 30 is an example of a confidence determination means. The lower limit value of the confidence is a criterion when the search unit 10 executes a search.

[0053] In the first embodiment, the lower limit value of the confidence was a preset fixed value. On the other hand, in the second embodiment, the lower limit value of the confidence is determined based on a user operation. In the second embodiment, the initial value of the lower limit value of the confidence is preset. However, the lower limit value of the confidence is variable.

[0054] A specific example of enabling the user to perform an operation (user operation) to change the lower limit value of the confidence from the search screen will be described later. The confidence determination unit 30 transmits the information on the lower limit value of the confidence determined based on the user operation to the search unit 10.

[0055] The search unit 10 performs a search for the target using the lower limit value of the confidence. More specifically, the search unit 10 searches for a target such that the value representing the likelihood of satisfying the search condition always exceeds the lower limit value of the confidence. For example, when the search condition is "male" and the lower limit value of the confidence is 80%, only the targets whose probability of the target attribute being "male" exceeds 80% are included in the search result by the search unit 10.

[0056] In addition, the confidence determination unit 30 presents the lower limit value of the confidence determined for each search condition. For example, the confidence determination unit 30 presents the lower limit value of the confidence determined for each search condition on a slider bar.

[0057] The confidence determination unit 30 accepts a user operation for changing the current lower limit value of the confidence for each search condition. For example, the confidence determination unit 30 accepts a user operation to move the slider on the slider bar. In this case, the confidence determination unit 30 determines the lower limit value of the confidence based on the position of the slider on the slider bar. An example of presenting the lower limit value of the confidence by the confidence determination unit 30 will be described later.

[0058] (Operation Flow of Information Processing Apparatus 2) With reference to FIG. 8, the operation flow of the information processing apparatus 2 according to Embodiment 1 will be described. FIG. 8 is a flowchart showing the flow of processes executed by each part of the information processing apparatus 2.

[0059] As shown in FIG. 8, the confidence determination unit 30 determines the lower limit value of the confidence (S101). When the user performs a user operation to change the lower limit value of the confidence, the lower limit value of the confidence is freely selected by the user. On the other hand, when the user has never changed the lower limit value of the confidence, the lower limit value of the confidence is a preset initial setting value. The confidence determination unit 30 transmits the information on the determined lower limit value of the confidence to the search unit 10.

[0060] The search unit 10 receives the information on the lower limit value of the confidence from the confidence determination unit 30. The search unit 10 uses the lower limit value of the confidence to search for a target that satisfies at least one of one or more search conditions (S102). The search unit 10 transmits the search result to the presentation unit 20.

[0061] As described in Embodiment 1, the search result includes at least one or more image data that satisfy at least one of the input one or more search conditions and information for specifying each target.

[0062] The presentation unit 20 receives the search result from the search unit 10. The presentation unit 20 presents in a list which search conditions the target satisfies and which search conditions it does not satisfy (S103). For example, as described in Embodiment 1, the presentation unit 20 displays a chart including a bar graph assigned to each search condition on the display (see, for example, FIG. 4 of Embodiment 1).

[0063] The confidence determination unit 30 determines whether a user operation for changing the lower limit value of the confidence has been performed (S104).

[0064] If a user operation for changing the lower limit value of the confidence has not been performed (No in S104), the operation of the information processing apparatus 2 ends.

[0065] On the other hand, when a user operation for changing the lower limit value of the confidence level is performed (Yes in S104), the flow returns to step S101. In step S101, the confidence level determination unit 30 adjusts the lower limit value of the confidence level based on the user operation. Then, the flow proceeds to step S102 described above.

[0066] (Presentation of the lower limit value of the confidence level; an example of a search screen) With reference to FIG. 9, an example of the presentation of the lower limit value of the confidence level by the confidence level determination unit 30 will be described. Also, an example of a user operation for changing the lower limit value of the confidence level will be described. FIG. 9 shows an example of a search screen according to the second embodiment. In the search screen shown in FIG. 9, for each attribute, the current lower limit value of the confidence level is shown as a numerical value (60%, 60%, 80%, 90%).

[0067] Also, a slider is displayed below the numerical value representing the current lower limit value of the confidence level. The left end of the slider corresponds to the lowest confidence level (0%), and the right end of the slider corresponds to the highest confidence level (100%). The user performs a user operation for changing the lower limit value of the confidence level. Specifically, the user changes the lower limit value of the confidence level by moving the slider on the slide bar. In FIG. 9, the triangular figure on the slide bar is the slider that is moved by the user.

[0068] The confidence level determination unit 30 determines the current lower limit value of the confidence level according to the amount of movement of the slider. In FIG. 9, the slider is triangular, but the shape and form of the slider are not particularly limited. The confidence level determination unit 30 transmits the information on the determined lower limit value of the confidence level to the search unit 10.

[0069] When the search unit 10 receives the information on the lower limit value of the confidence level determined by the confidence level determination unit 30, or when the search key is pressed or clicked, the search unit 10 re-searches for the target using the lower limit value of the confidence level determined by the confidence level determination unit 30.

[0070] (Effects of this embodiment) According to the configuration of the present embodiment, the search unit 10 searches for a target that satisfies one or more input search conditions, and the presentation unit 20 presents in a list which search conditions among the one or more search conditions the target satisfies and which search conditions the target does not satisfy based on the search result by the search unit 10. When a plurality of search conditions are input, the user can understand the target that satisfies two or more search conditions that are all or part of the plurality of presented search conditions by checking the presented list. Therefore, it is possible to help the user know the target that satisfies a plurality of search conditions.

[0071] Furthermore, the confidence determination unit 30 determines a lower limit value of the confidence, which is a value representing the certainty of the search result. The search unit 10 executes the search for the target using the lower limit value of the confidence. Therefore, it is possible to obtain as a search result a target whose value representing the certainty of satisfying the search condition exceeds the lower limit value of the confidence.

[0072] 〔Embodiment 3〕 Referring to FIGS. 10 to 12, Embodiment 3 will be described.

[0073] (Information processing apparatus 3) FIG. 10 is a block diagram showing the configuration of the information processing apparatus 3 according to Embodiment 3. As shown in FIG. 10, the information processing apparatus 3 includes a search unit 10, a presentation unit 20, and a recommendation unit 40. That is, the information processing apparatus 3 according to Embodiment 3 is different in configuration from the information processing apparatus 1 according to the first embodiment in that it includes a recommendation unit 40.

[0074] The recommendation unit 40 recommends other search conditions for further narrowing down the targets included in the search result. The recommendation unit 40 is an example of a recommendation means.

[0075] Specifically, the recommendation unit 40 receives, from the search unit 10, the results of a search based on one or more other search condition candidates (hereinafter sometimes referred to as second search conditions) that are different from the one or more input search conditions (hereinafter sometimes referred to as first search conditions). The search results include information indicating whether the target satisfies the other search condition candidates (second search conditions) or not. The search results also include information for identifying the targets included in the search results.

[0076] Based on the received search results, the recommendation unit 40 selects, from among the one or more other search condition candidates (second search conditions), the other search conditions to be recommended (hereinafter sometimes referred to as third search conditions).

[0077] Specifically, the recommendation unit 40 preferably selects, as the other search conditions to be recommended (third search conditions), other search condition candidates (second search conditions) such that the targets can be divided as evenly as possible.

[0078] That is, the recommendation unit 40 preferably determines the other search conditions to be recommended (third search conditions) such that the targets satisfying the one or more input search conditions (first search conditions) are divided into approximately 50% that satisfy the other search conditions and approximately 50% that do not satisfy the other search conditions, or as close as possible to a ratio of 50%:50%. One of the reasons is to reduce the bias between the cost (the effort of the user) of investigating all the targets that satisfy the other search conditions and the cost of investigating all the targets that do not satisfy the other search conditions when dividing the investigation of the targets among two or more users.

[0079] The recommendation unit 40 recommends the other search conditions determined in this way to the user. For example, the recommendation unit 40 presents a message or the like prompting the addition of the other search conditions.

[0080] In addition, the recommendation unit 40 presents the search results when searching for a target under other search conditions. For example, the recommendation unit 40 presents the search results obtained using other search conditions in a pie chart. An example of the presentation by the recommendation unit 40 will be described later.

[0081] (Operation flow of information processing apparatus 3) With reference to FIG. 11, the operation flow of the information processing apparatus 3 according to Embodiment 3 will be described. FIG. 11 is a flowchart showing the processing flow executed by each part of the information processing apparatus 3.

[0082] As shown in FIG. 11, the search unit 10 searches for a target that satisfies at least one of the one or more input search conditions (first search conditions) using the lower limit value of the confidence level (S201). Also in Embodiment 3, similar to Embodiment 1, the lower limit value of the confidence level is fixed. However, in a modified example, the information processing apparatus 3 may include the confidence level determination unit 30 described in Embodiment 2. In this case, the lower limit value of the confidence level is variable, and based on a user operation, the lower limit value of the confidence level is determined.

[0083] The search unit 10 transmits the search results obtained using the first search condition to the presentation unit 20. As described in Embodiment 1, the search results include information indicating which search conditions each target satisfies or does not satisfy, and information for identifying each target.

[0084] Based on the search results by the search unit 10, the presentation unit 20 presents in a list which search conditions the target satisfies and which search conditions it does not satisfy among the one or more search conditions (S202). For example, the presentation unit 20 displays a chart including bar graphs assigned to each search condition on the display, as in Embodiment 1 (see FIG. 4).

[0085] The recommendation unit 40 acquires the search results using one or more other search condition candidates (second search conditions) from the search unit 10 (S203).

[0086] The recommendation unit 40 determines other search conditions (third search conditions) to be recommended based on the results of searches using one or more other search condition candidates (second search conditions) (S204). As described above, the recommendation unit 40 preferably selects, as other search conditions, candidates for other search conditions (second search conditions) that can divide the target as evenly as possible.

[0087] The recommendation unit 40 presents the results of the search using the selected other search conditions (S205).

[0088] Thus, the operation of the information processing apparatus 3 ends. Note that the order of steps S201 to S202 and steps S203 to S205 is not limited. That is, steps S203 to S205 may be executed earlier than steps S201 to S202.

[0089] (Recommendation of Other Search Conditions) FIG. 12 shows an example of a presentation for the recommendation unit 40 to recommend other search conditions. As described above, the other search conditions to be recommended (third search conditions) are determined by the recommendation unit 40. In the example shown in FIG. 12, "necktie" corresponds to the other search conditions to be recommended (third search conditions).

[0090] FIG. 12 shows, in a pie chart, the search results when a search is executed using the other search conditions to be recommended (i.e., "necktie"). However, the search results when a search is executed using the other search conditions to be recommended (third search conditions) may be presented by means other than a pie chart.

[0091] According to FIG. 12, when "necktie" is added to the one or more input search conditions (first search conditions), the target is divided into 62.5%:37.5% depending on the presence or absence of a necktie. As described above, this ratio is preferably closer to 50%:50%.

[0092] (Effects of This Embodiment) According to the configuration of this embodiment, the search unit 10 searches for a target that satisfies one or more input search conditions, and the presentation unit 20 presents, in a list, which search conditions among the one or more search conditions the target satisfies and which search conditions the target does not satisfy based on the result of the search by the search unit 10. When a plurality of search conditions are input, the user can understand a target that satisfies two or more search conditions that are all or a part of the plurality of presented search conditions by checking the presented list. Therefore, it is possible to help the user know a target that satisfies a plurality of search conditions.

[0093] Furthermore, the recommendation unit 40 recommends other search conditions for further narrowing down the targets included in the search result. Thereby, the user can further narrow down the targets included in the search result by combining other search conditions with the plurality of search conditions.

[0094] [Embodiment 4] Embodiment 4 will be described below with reference to FIG. 13.

[0095] (Regarding the hardware configuration) Each component of the information processing apparatus described in the above Embodiments 1 to 3 indicates a block of a functional unit. Some or all of these components are realized by, for example, an information processing apparatus 900 as shown in FIG. 13. FIG. 13 is a block diagram showing an example of the hardware configuration of the information processing apparatus 900.

[0096] As shown in FIG. 13, the information processing apparatus 900 includes, as an example, the following configuration.

[0097] ·CPU (Central Processing Unit) 901 ·ROM (Read Only Memory) 902 ·RAM (Random Access Memory) 903 ·Program 904 loaded into RAM 903 ·Storage device 905 that stores program 904 ·Drive device 907 that reads and writes recording medium 906 · A communication interface 908 connected to the communication network 909 · An input / output interface 910 for inputting and outputting data · A bus 911 connecting each component Each component of the information processing apparatus described in the first to third embodiments is realized by the CPU 901 reading and executing a program 904 that realizes these functions. The program 904 that realizes the functions of each component is stored in advance in, for example, the storage device 905 or the ROM 902, and is loaded into the RAM 903 by the CPU 901 and executed as needed. Note that the program 904 may be supplied to the CPU 901 via the communication network 909, or may be stored in advance in the recording medium 906, and the drive device 907 may read the program and supply it to the CPU 901.

[0098] (Effect of this embodiment) According to the configuration of this embodiment, the information processing apparatus described in the above embodiment is realized as hardware. Therefore, the same effects as those described in the above embodiment can be achieved.

[0099] The present invention has been described above by taking the above-described embodiments as exemplary examples. However, the present invention is not limited to the above-described embodiments. That is, various aspects that can be understood by those skilled in the art can be applied within the scope of the present invention in the above-described embodiments.

Explanation of reference numerals

[0100] 1 Information processing apparatus 2 Information processing apparatus 3 Information processing apparatus 900 Information processing apparatus 10 Search unit 20 Presentation unit 30 Confidence determination unit 40 Recommendation unit

Claims

1. Threshold determination means for determining a threshold for determining whether a person image satisfies each of a plurality of appearance attributes input by a user as search conditions; Calculation means for calculating a similarity between first feature data extracted from the person image and second feature data related to each of the plurality of appearance attributes; Search means for performing a search for a person image having a similarity exceeding the threshold determined for each of the plurality of appearance attributes, comprising: The threshold is changed by a user input, The search is executed when the threshold is changed, An information processing apparatus.

2. Storage means for storing in association with the person image and information indicating an attribute of a person reflected in the person image in a storage unit, comprising: The search means performs a search for the person image having a similarity exceeding the threshold determined for each of the plurality of appearance attributes from among the person images stored in the storage unit. The information processing apparatus according to claim 1.

3. The search means calculates the similarity between the first feature data extracted from the person image and the second feature data extracted from a sample image tagged with a keyword indicating the attribute input by the user. The information processing apparatus according to claim 1.

4. The threshold determination means presents the threshold determined for each search condition on a slide bar. The information processing apparatus according to claim 1.

5. Recommendation means for recommending other search conditions for further narrowing down the person images included in the search results, Further comprising: The information processing apparatus according to any one of claims 1 to 4.

6. The recommendation means, Presents the result of the search by the search means when the other search conditions are used. The information processing apparatus according to claim 5.

7. A computer, Determines a threshold for determining whether a person image satisfies each of a plurality of appearance attributes input by a user as search conditions, Calculates a similarity between first feature data extracted from the person image and second feature data related to each of the plurality of appearance attributes, Performs a search for a person image having a similarity exceeding the threshold determined for each of the plurality of appearance attributes, which is an information processing method, The threshold is changed by a user input, The search is executed when the threshold is changed, An information processing method.

8. A program for causing a computer to perform: a process of determining a threshold value for determining whether a person image satisfies each of a plurality of appearance attributes input by a user as search conditions; a process of calculating a similarity between first feature data extracted from the person image and second feature data related to each of the plurality of appearance attributes; a process of executing a search for a person image having a similarity exceeding the threshold value determined for each of the plurality of appearance attributes, wherein the threshold value is changed by user input, and the search is executed when the threshold value is changed. Program.

Images