# US 11,308,352 B2 · Apparatus for processing labeled data to be used in learning of discriminator, method of controlling the apparatus, and non-transitory computer-readable recording medium > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,308,352 B2 - Document ID: US11308352B2 - Status: Granted - Kind: B2 - Application number: 16/798,602 - Filed: 2020-02-24 - Published: 2020-09-03 - Granted: 2022-04-19 - Earliest claimed: 2019-02-28 - Assignees: Canon Kabushiki Kaisha - Applicants: Canon Kabushiki Kaisha - Inventors: Atsushi Nogami; Masakazu Matsugu - Web page: https://patentgrep.com/patent/US11308352B2 ## CPC - G06V 10/235: {based on user input or interaction} - G06F 18/2132: based on discrimination criteria, e.g. discriminant analysis - G06F 18/214: Generating training patterns; Bootstrap methods, e.g. bagging or boosting - G06F 18/2178: {based on feedback of a supervisor} - G06F 18/23: Clustering techniques - G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting - G06V 10/7788: {the supervisor being a human, e.g. interactive learning with a human teacher} - G06V 20/20: in augmented reality scenes - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract An apparatus comprising: an obtaining unit configured to obtain target data as a result of discrimination of each portion of input data performed by a discriminator having learned in advance by using existing labeled data; a setting unit configured to set each portion of the target data, which is effective for additional learning of the discriminator, as local data; a determining unit configured to determine not less than one partial region of the target data, which accepts labeling by a user, based on a distribution of the set local data in the target data; and a display control unit configured to cause a display unit to display the determined not less than one partial region. ## Claims ### Claim 1 An apparatus comprising: one or more processors; and at least one medium storing instructions to be executed by the one or more processors, wherein the instructions when executed cause the apparatus to: obtain target data as a result of discrimination of each portion of input data performed by a discriminator having learned in advance by using existing labeled data; set each portion of the target data, which is effective for additional learning of the discriminator, as local data; determine not less than one partial region of the target data, which accepts labeling by a user, based on a distribution of the set local data in the target data; and display on a display device the determined not less than one partial region. ### Claim 2 (depends on 1) The apparatus according to claim 1, wherein the the instructions when executed further cause the apparatus to narrow down a portion of the target data, which accepts labeling by the user, as the partial region, based on the distribution of the not less than one local data in the target data. ### Claim 3 (depends on 1) The apparatus according to claim 1, wherein the partial region contains a plurality of local data. ### Claim 4 (depends on 1) The apparatus according to claim 1, wherein the not less than one portion of the target data, which is effective for additional learning of the discriminator, is a portion having a feature different from a feature learned by the existing labeled data. ### Claim 5 (depends on 1) The apparatus according to claim 1, wherein the local data is set based on similarity or divergence between the target data and a data set of the existing labeled data. ### Claim 6 (depends on 1) The apparatus according to claim 1, wherein the partial region is determined based on a density of the local data. ### Claim 7 (depends on 1) The apparatus according to claim 1, wherein the distribution of the local data is clustered, and the partial region is determined based on the clustering result. ### Claim 8 (depends on 1) The apparatus according to claim 1, wherein a plurality of partial region candidates are formed based on the distribution of the local data, priority of each of the plurality of partial region candidates is calculated, and the partial region from the partial region candidates is selected based on the priority. ### Claim 9 (depends on 8) The apparatus according to claim 8, wherein the the priority of each of the partial region candidates is calculated based on a density of the local data contained in the partial region candidates. ### Claim 10 (depends on 8) The apparatus according to claim 8, wherein importance of each local data is calculated, and the priority of each of the partial region candidates is calculated based on the importance of the local data contained in the partial region candidates. ### Claim 11 (depends on 1) The apparatus according to claim 1, wherein a plurality of partial region candidates are formed based on the distribution of the local data, and the partial region is selected from the partial region candidates based on a positional relationship of each partial region candidate. ### Claim 12 (depends on 1) The apparatus according to claim 1, wherein priority of the partial region is calculated based on the local data contained in the partial region, and a display order of the partial regions is controlled based on the priority of each partial region. ### Claim 13 (depends on 1) The apparatus according to claim 1, wherein a display order of the partial regions is controlled based on a positional relationship of each partial region. ### Claim 14 (depends on 1) The apparatus according to claim 1, wherein the display device displays a window including a GUI for accepting labeling performed on the partial region by the user. ### Claim 15 (depends on 14) The apparatus according to claim 14, the instructions when executed further cause the apparatus to accept input of labeling, and wherein, if input of labeling exceeding a range of the partial region is accepted, the range of the partial region is extended based on a range of the input labeling. ### Claim 16 (depends on 1) The apparatus according to claim 1, wherein the input data is image data obtained by imaging a structure, and wherein the discriminator discriminates between a portion showing a predetermined defect and a portion not showing the predetermined defect in the image data. ### Claim 17 (depends on 16) The apparatus according to claim 16, wherein the predetermined defect is a crack generated in the structure, and wherein the labeling is a result of clearly showing a pixel indicating a crack in the partial region displayed on the display device. ### Claim 18 An information processing method to be executed by an apparatus, the method comprising: obtaining target data as a result of discrimination of each portion of input data performed by a discriminator having learned in advance by using existing labeled data; setting each portion of the target data, which is effective for additional learning of the discriminator, as local data; determining not less than one partial region of the target data, which accepts labeling by a user, based on a distribution of the set local data in the target data; and causing a display device to display the determined not less than one partial region. ### Claim 19 A non-transitory computer-readable recording medium storing a program that, when executed by a computer, causes the computer to: obtain target data as a result of discrimination of each portion of input data performed by a discriminator having learned in advance by using existing labeled data; set each portion of the target data, which is effective for additional learning of the discriminator, as local data; determine not less than one partial region of the target data, which accepts labeling by a user, based on a distribution of the set local data in the target data; and display on a display device the determined not less than one partial region.