Agency for Science, Technology and Research

Nanyang Technological University

Nbcuniversal Media, LLC

Comcast Cable Communications, LLC

Wei Yun Yau

Xudong Jiang

Sheau Bao Ng

Wee Ser

Sujoy Roy

Kong Wah Wan

Tai Pang Chen

Oliver Jojic

Amit Bagga

Zhengguo Li

Jinghong Zheng

Hassan Sayyadi-Harikandehei

Zijian Orson Zhu

How Lung Eng

Zhiyong Huang

Ghim Eng Yap

Junxian Wang

Siew Wah Alvin Harvey Kam

Kar Ann Toh

Boon Hwa Tan

Qinyuan Ren

Renjun Li

Xianjun Wang

Qing'An Li

Shuguo Han

A camera-based localization system is provided. The camera-based localization system may assist an unmanned vehicle to continue its operation in a GPS-denied environment with minimal increase in vehicular cost and payload. In one aspect, a method, a computer-readable medium, and an apparatus for localization via visual inertial odometry are provided. The apparatus may construct an optical flow based on feature points across a first video frame and a second video frame captured by a camera of the apparatus. The apparatus may refine the angular velocity and the linear velocity corresponding to the second video frame via solving a quadratic optimization problem constructed based on the optical flow, the initial values of the angular velocity and the linear velocity corresponding to the second video frame. The apparatus may estimate the pose of the apparatus based on the refined angular velocity and the refined linear velocity.

Real time robust localization via visual inertial odometry

There is provided a method of controlling a mobile manipulator for tracking a surface. The mobile manipulator includes a mobile base movable in an axial direction of the mobile manipulator and a manipulator supported on the mobile base having an end effector adjustable in a lateral direction of the mobile manipulator. The method includes detecting the surface from the mobile manipulator, including positions of the surface at points along the surface, determining a reference path for the end effector to track based on an offset from the surface detected, determining a tracking error in the reference path determined, and adjusting a position of the end effector in the lateral direction based on the tracking error to compensate for the tracking error in the reference path determined. There is also provided a corresponding mobile manipulator.

Mobile manipulator and method of controlling the mobile manipulator for tracking a surface

Systems and methods for transforming a similarity matrix used to generate content item recommendation lists are described. The system may determine a weight (used to transform the similarity matrix) sufficient to place one or more content items on the content item recommendation list or at a particular position on the recommendation list. The transformed similarity matrix may further be reduced in size, and content recommendations may be generated using the transformed or reduced similarity matrix.

Recommendation system using a transformed similarity matrix

A method and system for modeling the interests of an individual television viewer based on logging the user's television usage activities for the purpose of determining personalized recommendations is presented. According to one aspect of the claimed subject matter, user input in response to a questionnaire, the user's recorded television usage, the user's previous interactions with the personalized recommendations and contextual data such as external events and social group influences are used to model the probability of a user's interest in available media content. Data from the model is subsequently arranged to display units of available media content with the highest probabilities of user interest to the user. According to further embodiments, specific user behavior in previously consumed media content such as a user's selection, a user's continuous consumption, and a user's discontinuation of consumption is tracked and used as input.

Method and system for personalized recommendation modeling

Various embodiments provide a method for collecting audience information of a group of users in an audience measurement. The method may include each user dividing its private information into a private portion and a predetermined number of exchange portions; each user transmitting each of its exchange portions to a respective user of the predetermined number of users selected from the group of users; each user aggregating its private portion and exchange portions received from other users of the group of users, thereby producing randomized information; and aggregating the randomized information of the group of users, thereby producing aggregated private information for the group of users.

Method and device for collecting audience information

Identity authentication systems and techniques are disclosed which solves the problem associated with limited processing power and smart card technology in the handling of biometric authentication. By distributing the processing of an identity authenticating process between a smart card and a computer terminal, the complicated calculation involved in a biometrics matching process can be carried out to allow verification using biometric parameters stored on smart cards. There is disclosed a system and technique for user authentication, together with a system and technique for distributed processing. A registration method is also described.

Distributed processing in authentication

Embodiment of the invention provide a method and a device for determining a similarity value between a first template and a second template. A first cluster characteristic for each first cluster of a plurality of first clusters is determined, wherein each first cluster includes a plurality of first minutiae comprised in the first minutiae template. A second cluster characteristic for each second cluster of a plurality of second clusters is determined, wherein each second cluster includes a plurality of second minutiae comprised in the second minutiae template. The similarity value between the first minutiae template and the second minutiae template is determined based on the first cluster characteristics and the second cluster characteristics.

Method and device for determining a similarity value between minutiae templates

A background image is generated based on a captured image. A data cluster is formed in pixel blocks of the background image using at least one feature of pixels in the pixel blocks. A data cluster formed in each pixel block includes a data distribution having a mean value and a standard deviation from the mean value. After generating the background image, each pixel of a subsequent captured image is compared with the data cluster of a pixel block of the background image to generate a first discrepancy value. A pixel of a subsequent image is compared with a data distribution of another adjacent pixel block of the background image to generate a second discrepancy value. Based on the discrepancy values, the pixels of the subsequent image are regarded as background or foreground pixels in a binary map, in which connected foreground pixels are marked to form a foreground object.

Method for detecting desired objects in a highly dynamic environment by a monitoring system

A template minutia set of a fingerprint comprising template minutiae and a template interest region is provided. Further, an input minutia set comprising input minutiae and an input interest region according to an input fingerprint is provided. The input minutia set is matched with the template minutia set. All matching template/input minutiae are determined. Furthermore, an updated template minutia set is determined dependent on the matching template/input minutiae and the non-matching template/input minutiae, and dependent on whether the template minutiae are inside the input interest region or outside of it and whether the input minutiae are inside the template interest region or outside of it.

Method and device for computer-based processing a template minutia set of a fingerprint and a computer readable storage medium

A reference minutiae template comprising reference minutiae and a first partial minutiae template, which overlaps the reference minutiae template, the first partial minutiae template comprising first minutiae, are selected. The reference minutiae of the reference minutiae template are matched with the first minutiae of the first partial minutiae template. Using the matched minutiae, a transformation function is determined, which describes the relation between the reference minutiae and the first minutiae. The first minutiae, which are located outside the overlapping region, are transformed using the transformation function, thereby generating a new reference template. Furthermore, a second partial minutiae template is selected, which overlaps the new reference minutiae template, the second partial minutiae template comprising second minutiae. The method is repeated iteratively for all partial minutiae templates with the new reference minutiae template as the reference minutiae template of the next iteration, the second partial minutiae template as the first partial minutiae template of the next iteration, thereby generating the total minutiae template.

Method and device for determining a total minutiae template from a plurality of partial minutiae templates

Two-stage local and global fingerprint matching technique for automated fingerprint verification/identification

Growing community of inventors

Singapore, Singapore

Wei Yun Yau