Patents by Inventor Kazuhiro Fukui
Kazuhiro Fukui has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20240029434Abstract: A non-transitory computer-readable recording medium has stored therein a program that causes a computer to execute a process, the process including acquiring movie data including a plurality of consecutive frames calculating first likelihood of a class of the movie data by inputting the acquired movie data to a trained model, calculating an optical flow indicating movement of an area included in the movie data, based on the movie data generating occluded movie data by setting an occluded area in each of the frames included in the movie data, based on the optical flow, calculating second likelihood of a class of the occluded movie data by inputting the occluded movie data to the model identifying an area that affects identification of the class among areas in the movie data, based on the first likelihood and the second likelihood and displaying the identified area that affects identification of the class.Type: ApplicationFiled: July 19, 2022Publication date: January 25, 2024Applicants: FUJITSU LIMITED, University of TsukubaInventors: Tomoki UCHIYAMA, Naoya SOGI, Koichiro NIINUMA, Kazuhiro FUKUI
-
Publication number: 20230316700Abstract: An information processing apparatus acquires an image; generates a shielding image in which a part of an area included in an area of the acquired image has been shielded; calculates, by inputting the image to a first model that has been trained, first likelihood of the target object included in the image; calculates, by inputting the shielding image to a second model that calculates an approximation value of likelihood of the target object included in the image when the image is input, second likelihood corresponding to the approximation value of the likelihood of the target object included in the shielding image; specifies, based on the first likelihood and the second likelihood, an area that affects discrimination of the class and that is included in the area of the image; and displays the specified area that affects discrimination of the class.Type: ApplicationFiled: March 31, 2022Publication date: October 5, 2023Applicants: FUJITSU LIMITED, University of TsukubaInventors: Naoya SOGI, Tomoki UCHIYAMA, Kazuhiro FUKUI, Koichiro NIINUMA
-
Publication number: 20220396981Abstract: The present application improves operability. In the door opening/closing detection apparatus, the handle is axially supported rotatably on the base with the axis direction in vertical direction. When the handle's gripping portion is rotated from the neutral position to the push operation position rotated toward the outside of the vehicle, the first switch is activated by the handle's switch operation portion. Meanwhile, when the handle's gripping portion is rotated from the neutral position to the pull operation position rotated toward the interior of the vehicle, the handle's cam surface activates the second switch. This allows detection of operations on the handle while keeping the direction of push and pull operations on the gripping portion of the handle in line with the opening/closing direction of the door D. This allows the passenger to intuitively recognize the operation to the handle when opening and closing the door D, leading to improved operability.Type: ApplicationFiled: October 5, 2020Publication date: December 15, 2022Inventors: Soichi MIYATA, Kazuhiro FUKUI
-
Patent number: 7454062Abstract: In one embodiment of the invention, a pattern recognition apparatus comprises a unit for inputting a pattern of a to-be recognized category; and a processor with a memory for: generating input subspace; calculating and storing reference subspaces; storing constraint subspaces for extracting features; projecting the input subspace and the reference subspaces respectively onto the constraint subspaces; calculating similarities between the respective reference subspaces and the input subspace in such projected state; combining the similarities in respect of the constraint subspaces on each of the reference subspaces; and identifying the to-be recognized category with a category corresponding to one of the reference subspaces, if the combined similarity between the one of reference subspace and the input subspace is highest among them.Type: GrantFiled: November 4, 2004Date of Patent: November 18, 2008Assignee: Kabushiki Kaisha ToshibaInventors: Masashi Nishiyama, Osamu Yamaguchi, Kazuhiro Fukui
-
Patent number: 7346192Abstract: An image input unit inputs an image including a user's face area to specify the user's gaze or face direction. A pattern generation unit generates an input pattern that can be compared with each dictionary pattern from the image. A dictionary unit previously stores a plurality of dictionary patterns each of which differently corresponds to the user's gaze or face direction. A pattern comparison unit respectively calculates a similarity degree between the input pattern and each of the dictionary patterns by comparing the input pattern with each of the dictionary patterns. A gaze detection unit detects the user's gaze or face direction based on the similarity degree.Type: GrantFiled: January 19, 2006Date of Patent: March 18, 2008Assignee: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kazuhiro Fukui, Osamu Yamaguchi
-
Patent number: 7330591Abstract: The present invention solves an eigenvalue problem using a sum of projection matrixes for each subspace. A space which is spanned by a plurality of eigenvectors which are selected starting from those having a smaller eigenvalue is used as a constraint subspace.Type: GrantFiled: February 25, 2003Date of Patent: February 12, 2008Assignee: Kabushiki Kaisha ToshibaInventors: Kazuhiro Fukui, Osamu Yamaguchi
-
Patent number: 7224835Abstract: An object recognizing apparatus can execute accurate object recognition based on dictionaries each created from a predetermined number of patterns, by computing a similarity between a subspace in each dictionary and a subspace created from input images. In this apparatus, images of a to-be-identified object obtained by an image input unit is expressed in the form of an input subspace by an inter-subspace-angle computing unit. The angles between the input subspace and the subspace of each dictionary stored in a dictionary storing unit is computed. An identifying unit identifies a to-be-identified object based on the angles (the degrees of similarity) computed by the inter-subspace-angle computing unit.Type: GrantFiled: February 24, 2003Date of Patent: May 29, 2007Assignee: Kabushiki Kaisha ToshibaInventors: Kenichi Maeda, Osamu Yamaguchi, Kazuhiro Fukui
-
Patent number: 7127086Abstract: An image processing apparatus. A camera inputs an image of a face of a person to be recognized. A recognition area detection section generates a difference image between the input image and a predetermined pattern, and detects a recognition area whose value is above a threshold from the input image. An input data generation section converts the recognition area to a predetermined input data. A similarity calculation section calculates a similarity by comparing the predetermined input data with a predetermined dictionary data. A view position of the camera is located lower than a position of the face of the person. A direction of the optical axis of the camera represents an angle of elevation from a horizontal direction.Type: GrantFiled: August 27, 2003Date of Patent: October 24, 2006Assignee: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kaoru Suzuki, Kazuhiro Fukui, Osamu Yamaguchi, Tomokazu Wakasugi
-
Patent number: 7065230Abstract: An image input unit inputs an image including a user's face area to specify the user's gaze or face direction. A pattern generation unit generates an input pattern that can be compared with each dictionary pattern from the image. A dictionary unit previously stores a plurality of dictionary patterns each of which differently corresponds to the user's gaze or face direction. A pattern comparison unit respectively calculates a similarity degree between the input pattern and each of the dictionary patterns by comparing the input pattern with each of the dictionary patterns. A gaze detection unit detects the user's gaze or face direction based on the similarity degree.Type: GrantFiled: May 24, 2002Date of Patent: June 20, 2006Assignee: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kazuhiro Fukui, Osamu Yamaguchi
-
Publication number: 20060115125Abstract: An image input unit inputs an image including a user's face area to specify the user's gaze or face direction. A pattern generation unit generates an input pattern that can be compared with each dictionary pattern from the image. A dictionary unit previously stores a plurality of dictionary patterns each of which differently corresponds to the user's gaze or face direction. A pattern comparison unit respectively calculates a similarity degree between the input pattern and each of the dictionary patterns by comparing the input pattern with each of the dictionary patterns. A gaze detection unit detects the user's gaze or face direction based on the similarity degree.Type: ApplicationFiled: January 19, 2006Publication date: June 1, 2006Inventors: Mayumi Yuasa, Kazuhiro Fukui, Osamu Yamaguchi
-
Publication number: 20050141767Abstract: In one embodiment of the invention, a pattern recognition apparatus comprises a unit for inputting a pattern of a to-be recognized category; and a processor with a memory for: generating input subspace; calculating and storing reference subspaces; storing constraint subspaces for extracting features; projecting the input subspace and the reference subspaces respectively onto the constraint subspaces; calculating similarities between the respective reference subspaces and the input subspace in such projected state; combining the similarities in respect of the constraint subspaces on each of the reference subspaces; and identifying the to-be recognized category with a category corresponding to one of the reference subspaces, if the combined similarity between the one of reference subspace and the input subspace is highest among them.Type: ApplicationFiled: November 4, 2004Publication date: June 30, 2005Applicant: KABUSHIKI KAISHA TOSHIBAInventors: Masashi Nishiyama, Osamu Yamaguchi, Kazuhiro Fukui
-
Publication number: 20040042644Abstract: An image processing apparatus. A camera inputs an image of a face of a person to be recognized. A recognition area detection section generates a difference image between the input image and a predetermined pattern, and detects a recognition area whose value is above a threshold from the input image. An input data generation section converts the recognition area to a predetermined input data. A similarity calculation section calculates a similarity by comparing the predetermined input data with a predetermined dictionary data. A view position of the camera is located lower than a position of the face of the person. A direction of the optical axis of the camera represents an angle of elevation from a horizontal direction.Type: ApplicationFiled: August 27, 2003Publication date: March 4, 2004Applicant: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kaoru Suzuki, Kazuhiro Fukui, Osamu Yamaguchi, Tomokazu Wakasugi
-
Patent number: 6697505Abstract: Configuration is made by a feature pattern extracting section 12 for extracting a feature obtainable from a person, a subspace calculating section 14 for calculating a subspace from the feature pattern, a recognition section 16 for effecting recognition using the subspace, a registered information holding section 18 for holding a plurality of the subspaces, and a registered information updating section 20 for updating the registered information using the subspaces. When identifying a person, the effect of aging change is reduced by suppressing time and year changes and keeping the recognition rate high.Type: GrantFiled: June 5, 2003Date of Patent: February 24, 2004Assignee: Kabushiki Kaisha ToshibaInventors: Osamu Yamaguchi, Kazuhiro Fukui
-
Patent number: 6690814Abstract: An image processing apparatus. A camera inputs an image of a face of a person to be recognized. A recognition area detection section generates a difference image between the input image and a predetermined pattern, and detects a recognition area whose value is above a threshold from the input image. An input data generation section converts the recognition area to a predetermined input data. A similarity calculation section calculates a similarity by comparing the predetermined input data with a predetermined dictionary data. A view position of the camera is located lower than a position of the face of the person. A direction of the optical axis of the camera represents an angle of elevation from a horizontal direction.Type: GrantFiled: March 10, 2000Date of Patent: February 10, 2004Assignee: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kaoru Suzuki, Kazuhiro Fukui, Osamu Yamaguchi, Tomokazu Wakasugi
-
Publication number: 20030198366Abstract: The present invention solves an eigenvalue problem using a sum of projection matrixes for each subspace. A space which is spanned by a plurality of eigenvectors which are selected starting from those having a smaller eigenvalue is used as a constraint subspace.Type: ApplicationFiled: February 25, 2003Publication date: October 23, 2003Inventors: Kazuhiro Fukui, Osamu Yamaguchi
-
Publication number: 20030194113Abstract: Configuration is made by a feature pattern extracting section 12 for extracting a feature obtainable from a person, a subspace calculating section 14 for calculating a subspace from the feature pattern, a recognition section 16 for effecting recognition using the subspace, a registered information holding section 18 for holding a plurality of the subspaces, and a registered information updating section 20 for updating the registered information using the subspaces. When identifying a person, the effect of aging change is reduced by suppressing time and year changes and keeping the recognition rate high.Type: ApplicationFiled: June 5, 2003Publication date: October 16, 2003Applicant: Kabushiki Kaisha ToshibaInventors: Osamu Yamaguchi, Kazuhiro Fukui
-
Publication number: 20030161537Abstract: An object recognizing apparatus can execute accurate object recognition based on dictionaries each created from a predetermined number of patterns, by computing a similarity between a subspace in each dictionary and a subspace created from input images. In this apparatus, images of a to-be-identified object obtained by an image input unit is expressed in the form of an input subspace by an inter-subspace-angle computing unit. The angles between the input subspace and the subspace of each dictionary stored in a dictionary storing unit is computed. An identifying unit identifies a to-be-identified object based on the angles (the degrees of similarity) computed by the inter-subspace-angle computing unit.Type: ApplicationFiled: February 24, 2003Publication date: August 28, 2003Inventors: Kenichi Maeda, Osamu Yamaguchi, Kazuhiro Fukui
-
Patent number: 6608914Abstract: Configuration is made by a feature pattern extracting section 12 for extracting a feature obtainable from a person, a subspace calculating section 14 for calculating a subspace from the feature pattern, a recognition section 16 for effecting recognition using the subspace, a registered information holding section 18 for holding a plurality of the subspaces, and a registered information updating section 20 for updating the registered information using the subspaces. When identifying a person, the effect of aging change is reduced by suppressing time and year changes and keeping the recognition rate high.Type: GrantFiled: December 11, 1998Date of Patent: August 19, 2003Assignee: Kabushiki Kaisha ToshibaInventors: Osamu Yamaguchi, Kazuhiro Fukui
-
Publication number: 20030039378Abstract: An image input unit inputs an image including a user's face area to specify the user's gaze or face direction. A pattern generation unit generates an input pattern that can be compared with each dictionary pattern from the image. A dictionary unit previously stores a plurality of dictionary patterns each of which differently corresponds to the user's gaze or face direction. A pattern comparison unit respectively calculates a similarity degree between the input pattern and each of the dictionary patterns by comparing the input pattern with each of the dictionary patterns. A gaze detection unit detects the user's gaze or face direction based on the similarity degree.Type: ApplicationFiled: May 24, 2002Publication date: February 27, 2003Applicant: Kabushiki Kaisha ToshibaInventors: Mayumi Yuasa, Kazuhiro Fukui, Osamu Yamaguchi
-
Patent number: RE39550Abstract: A security apparatus. A service is supplied to a user while maintaining the security of the service. A person discrimination section discriminates the user to be supplied the service. A user situation decision section decides whether the user is under a situation to use the service. An infringement situation decision section detects whether a non-user intrudes into a use area of the service in order to decide whether the security of the service is infringed. A service control section supplies the service to the user in case the person discrimination section discriminates the user, and controls a supply of the service if the use situation decision section decides the user is not under the situation to use the service or the infringement situation decision section decides that the security of the service is infringed.Type: GrantFiled: June 13, 2001Date of Patent: April 3, 2007Assignee: Kabushiki Kaisha ToshibaInventors: Kaoru Suzuki, Kazuhiro Fukui, Hisashi Kazama, Osamu Yamaguchi, Eiji Tanaka, Yasuhiro Taniguchi