Patents by Inventor Toufiq Parag

Toufiq Parag 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: 20240338839
    Abstract: Systems, apparatuses, and methods are described for detecting motion and/or objects in images using distance data. An image of an area to be monitored may be used to generate a distance map that maps distances between features in the area and a camera that took the image. Different reactions may be performed based on motion and/or objects detected in images from the camera and determined to be within different distance reaction zones relative to the camera. The different distance reaction zones and reactions may be based on customizable reaction rules. The distance reaction zones and/or reaction rules may be suggested based on similarities between the area to be monitored and other monitored areas.
    Type: Application
    Filed: April 7, 2023
    Publication date: October 10, 2024
    Inventors: Donald Tolley, Hongcheng Wang, Karen Chung, Sara Cuesta Gonzalez, Toufiq Parag
  • Publication number: 20240112125
    Abstract: Systems, apparatuses, and methods are described for detection of an item being delivered at a premises. Based on a movement pattern shown in a video captured by a camera associated with a premises, a delivery event may be determined. A notification of the delivery event may be sent to a user device associated with the premises.
    Type: Application
    Filed: October 3, 2022
    Publication date: April 4, 2024
    Inventors: Hongcheng Wang, Luke Deluccia, Sara Cuesta Gonzalez, Zhe Wu, Toufiq Parag
  • Patent number: 8548231
    Abstract: First order predicate logics are provided, extended with a bilattice based uncertainty handling formalism, as a means of formally encoding pattern grammars, to parse a set of image features, and detect the presence of different patterns of interest implemented on a processor. Information from different sources and uncertainties from detections, are integrated within the bilattice framework. Automated logical rule weight learning in the computer vision domain applies a rule weight optimization method which casts the instantiated inference tree as a knowledge-based neural network, to converge upon a set of rule weights that give optimal performance within the bilattice framework. Applications are in (a) detecting the presence of humans under partial occlusions and (b) detecting large complex man made structures in satellite imagery (c) detection of spatio-temporal human and vehicular activities in video and (c) parsing of Graphical User Interfaces.
    Type: Grant
    Filed: March 16, 2010
    Date of Patent: October 1, 2013
    Assignee: Siemens Corporation
    Inventors: Vinay Damodar Shet, Maneesh Kumar Singh, Claus Bahlmann, Visvanathan Ramesh, Stephen P. Masticola, Jan Neumann, Toufiq Parag, Michael A. Gall, Roberto Antonio Suarez
  • Patent number: 7840061
    Abstract: A method adapts a boosted classifier to new samples. A boosted classifier is trained using initial samples. The boosted classifier is a combination of weak classifiers. Each weak classifier of the boosted classifier is updated adaptively by adding contributions of new samples and deleting contributions old samples.
    Type: Grant
    Filed: February 28, 2007
    Date of Patent: November 23, 2010
    Assignee: Mitsubishi Electric Research Laboratories, Inc.
    Inventors: Fatih M. Porikli, Toufiq Parag
  • Publication number: 20100278420
    Abstract: First order predicate logics are provided, extended with a bilattice based uncertainty handling formalism, as a means of formally encoding pattern grmmars, to parse a set of image features, and detect the presence of different patterns of interest implemented on a processor. Information from different sources and uncertainties from detections, are integrated within the bilattice framework. Automated logical rule weight learning in the computer vision domain applies a rule weight optimization method which casts the instantiated inference tree as a knowledge-based neural network, to converge upon a set of rule weights that give optimal performance within the bilattice framework. Applications are in (a) detecting the presence of humans under partial occlusions and (b) detecting large complex man made structures in satellite imagery (c) detection of spatio-temporal human and vehicular activities in video and (c) parsing of Graphical User Interfaces.
    Type: Application
    Filed: March 16, 2010
    Publication date: November 4, 2010
    Applicant: Siemens Corporation
    Inventors: Vinay Damodar Shet, Maneesh Kumar Singh, Claus Bahlmann, Visvanathan Ramesh, Stephen P. Masticola, Jan Neumann, Toufiq Parag, Michael A. Gall, Roberto Antonio Suarez
  • Publication number: 20080205750
    Abstract: A method adapts a boosted classifier to new samples. A boosted classifier is trained using initial samples. The boosted classifier is a combination of weak classifiers. Each weak classifier of the boosted classifier is updated adaptively by adding contributions of new samples and deleting contributions old samples.
    Type: Application
    Filed: February 28, 2007
    Publication date: August 28, 2008
    Inventors: Fatih M. Porikli, Toufiq Parag