Patents by Inventor Vasudev Bhaskaran
Vasudev Bhaskaran 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).
-
Patent number: 11710344Abstract: A method is presented. The method includes determining a number of landmarks in an image comprising multiple pixels. The method also includes determining a number of channels for the image based on a function of the number of landmarks. The method further includes determining, for each one of the number of channels, a confidence of each pixel of the multiple pixels corresponding to a landmark. The method still further includes identifying the landmark in the image based on the confidence.Type: GrantFiled: April 27, 2020Date of Patent: July 25, 2023Assignee: QUALCOMM IncorporatedInventors: Upal Mahbub, Rakesh Nattoji Rajaram, Vasudev Bhaskaran
-
Publication number: 20220076059Abstract: Embodiments include systems and methods that may be performed by a processor of a computing device. Embodiments may be applied for keypoint detection in an image. In embodiments, the processor of the computing device may apply to an image a first-stage neural network to define and output a plurality of regions, apply to each of the plurality of regions a respective second-stage neural network to output a plurality of keypoints in each of the plurality of regions, and apply to the plurality of keypoints a third-stage neural network to determine a correction for each of the plurality of keypoints to provide corrected keypoints.Type: ApplicationFiled: November 19, 2021Publication date: March 10, 2022Inventors: Upal MAHBUB, Rakesh NATTOJI RAJARAM, Vasudev BHASKARAN
-
Patent number: 11256956Abstract: Embodiments include systems and methods for keypoint detection in an image. In embodiments, a processor of a computing device may apply to an image a first neural network that has been trained to define and output a plurality of regions. The processor may apply to each of the plurality of regions a respective second neural network to that has been trained to output a plurality of keypoints in each of the plurality of regions. The processor may apply to the plurality of keypoints a third neural network that has been trained to determine a correction for each of the plurality of keypoints to provide corrected keypoints suitable for the execution of an image processing function.Type: GrantFiled: December 2, 2019Date of Patent: February 22, 2022Assignee: Qualcomm IncorporatedInventors: Upal Mahbub, Rakesh Nattoji Rajaram, Vasudev Bhaskaran
-
Publication number: 20210334516Abstract: A method is presented. The method includes determining a number of landmarks in an image comprising multiple pixels. The method also includes determining a number of channels for the image based on a function of the number of landmarks. The method further includes determining, for each one of the number of channels, a confidence of each pixel of the multiple pixels corresponding to a landmark. The method still further includes identifying the landmark in the image based on the confidence.Type: ApplicationFiled: April 27, 2020Publication date: October 28, 2021Inventors: Upal MAHBUB, Rakesh NATTOJI RAJARAM, Vasudev BHASKARAN
-
Publication number: 20210166070Abstract: Embodiments include systems and methods for keypoint detection in an image. In embodiments, a processor of a computing device may apply to an image a first neural network that has been trained to define and output a plurality of regions. The processor may apply to each of the plurality of regions a respective second neural network to that has been trained to output a plurality of keypoints in each of the plurality of regions. The processor may apply to the plurality of keypoints a third neural network that has been trained to determine a correction for each of the plurality of keypoints to provide corrected keypoints suitable for the execution of an image processing function.Type: ApplicationFiled: December 2, 2019Publication date: June 3, 2021Inventors: Upal MAHBUB, RAKESH NATTOJI RAJARAM, Vasudev BHASKARAN
-
Publication number: 20190333229Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.Type: ApplicationFiled: July 11, 2019Publication date: October 31, 2019Inventors: Gokce DANE, Vasudev BHASKARAN
-
Publication number: 20190266429Abstract: A classifier for detecting objects in images can be configured to receive features of an image from a feature extractor. The classifier can determine a feature window based on the received features, and allows access by each decision tree of the classifier to only a predetermined area of the feature window. Each decision tree of the classifier can compare a corresponding predetermined area of the feature window with one or more thresholds. The classifier can determine an object in the image based on the comparisons. In some examples, the classifier can determine objects in a feature window based on received features, where the received features are based on color information for an image.Type: ApplicationFiled: February 23, 2018Publication date: August 29, 2019Inventors: Sujith Srinivasan, Rakesh Nattoji Rajaram, Gokce Dane, Vasudev Bhaskaran
-
Patent number: 10395377Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.Type: GrantFiled: January 16, 2018Date of Patent: August 27, 2019Assignee: QUALCOMM IncorporatedInventors: Gokce Dane, Vasudev Bhaskaran
-
Publication number: 20180158193Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.Type: ApplicationFiled: January 16, 2018Publication date: June 7, 2018Inventors: Gokce Dane, Vasudev Bhaskaran
-
Patent number: 9905015Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.Type: GrantFiled: August 24, 2017Date of Patent: February 27, 2018Assignee: QUALCOMM INCORPORATEDInventors: Gokce Dane, Vasudev Bhaskaran
-
Publication number: 20180012367Abstract: A method performed by an electronic device is described. The method includes generating a depth map of a scene external to a vehicle. The method also includes performing first processing in a first direction of a depth map to determine a first non-obstacle estimation of the scene. The method also includes performing second processing in a second direction of the depth map to determine a second non-obstacle estimation of the scene. The method further includes combining the first non-obstacle estimation and the second non-obstacle estimation to determine a non-obstacle map of the scene. The combining includes combining comprises selectively using a first reliability map of the first processing and/or a second reliability map of the second processing The method additionally includes navigating the vehicle using the non-obstacle map.Type: ApplicationFiled: August 24, 2017Publication date: January 11, 2018Inventors: Gokce Dane, Vasudev Bhaskaran
-
Publication number: 20170300742Abstract: An electronic device is described. The electronic device includes a memory configured to store a composite search space comprising a plurality of adjacent cells. Each of the adjacent cells includes a representation of an object. The electronic device also includes a dedicated engine configured to match a representation of an object from a captured image with the representations of the objects in the composite search space.Type: ApplicationFiled: April 14, 2016Publication date: October 19, 2017Inventors: Jian Wei, Ashwath Aithal, Vasudev Bhaskaran
-
Patent number: 9761000Abstract: A method performed by an electronic device is described. The method includes performing vertical processing of a depth map to determine a vertical non-obstacle estimation. The method also includes performing horizontal processing of the depth map to determine a horizontal non-obstacle estimation. The method further includes combining the vertical non-obstacle estimation and the horizontal non-obstacle estimation. The method additionally includes generating a non-obstacle map based on the combination of the vertical and horizontal non-obstacle estimations.Type: GrantFiled: September 18, 2015Date of Patent: September 12, 2017Assignee: QUALCOMM IncorporatedInventors: Gokce Dane, Vasudev Bhaskaran
-
Publication number: 20170084038Abstract: A method performed by an electronic device is described. The method includes performing vertical processing of a depth map to determine a vertical non-obstacle estimation. The method also includes performing horizontal processing of the depth map to determine a horizontal non-obstacle estimation. The method further includes combining the vertical non-obstacle estimation and the horizontal non-obstacle estimation. The method additionally includes generating a non-obstacle map based on the combination of the vertical and horizontal non-obstacle estimations.Type: ApplicationFiled: September 18, 2015Publication date: March 23, 2017Inventors: Gokce Dane, Vasudev Bhaskaran
-
Patent number: 9509970Abstract: In one example, a method includes identifying a pixel in an image frame that is a candidate for causing crosstalk between the image frame and a corresponding image frame in a multiview image system. The method further includes, for a pixel identified as a candidate for causing crosstalk, applying crosstalk correction to the pixel. The method further includes applying a location-based adjustment to the pixel, wherein the location-based adjustment is based at least in part on which of two or more portions of the image frame the pixel is in.Type: GrantFiled: July 18, 2012Date of Patent: November 29, 2016Assignee: QUALCOMM IncorporatedInventors: Gokce Dane, Vasudev Bhaskaran
-
Patent number: 9374506Abstract: Method and apparatus for reducing random noise in digital video streams are described. In one innovative aspect, the device includes a noise estimator. The device also includes a motion detector configured to determine a motion value indicative of motion between two frames of the video stream, the motion value based at least in part on the noise value. The device further includes a spatial noise reducer configured to filter the image data based at least in part on a blending factor and the noise value. The device also includes a temporal noise reducer configured to filter the video data based on the motion value and the noise value. The device also includes a blender configured to blend the spatial and temporal filtered values to provide a weighted composite filtered output image.Type: GrantFiled: January 4, 2013Date of Patent: June 21, 2016Assignee: QUALCOMM INCORPORATEDInventors: Mainak Biswas, Vasudev Bhaskaran, Sujith Srinivasan, Shilpi Sahu
-
Patent number: 9286658Abstract: Techniques described in the disclosure are generally related to enhancing portions of an image relative to other portions of the image. The example techniques may utilize depth information in conjunction with one or more viewer perceivable information to enhance portions of the image relative to other portions. The techniques may then display the enhanced image to provide the viewer with a possibly more realistic image.Type: GrantFiled: March 22, 2012Date of Patent: March 15, 2016Assignee: QUALCOMM IncorporatedInventor: Vasudev Bhaskaran
-
Patent number: 9232226Abstract: Systems and methods for compressing video data are provided. The method includes segmenting a video frame, selecting a coding mode, and encoding. The segmenting includes segmenting the video frame of the video data into a sequence of coding blocks. The selecting includes selecting the coding mode from a plurality of coding modes. The selecting of the coding mode is based on an allowable bit budget and occurs for each coding block. The encoding includes encoding each coding block based on the coding mode. The allowable bit budget varies according to a bit utilization of prior encoded coding blocks and varies such that the video frame does not exceed a specified compression ratio.Type: GrantFiled: August 17, 2009Date of Patent: January 5, 2016Assignee: MARVELL WORLD TRADE LTD.Inventors: Vasudev Bhaskaran, Nikhil Balram
-
Patent number: 9083948Abstract: In one example, a method includes identifying a first set of pixels in co-located pairs in a corresponding pair of multiview image frames for which the co-located pairs have a disparity between the pixels that is greater than a selected disparity threshold. The method further includes identifying a second set of pixels in at least one of the image frames that are within a selected distance of an intensity transition greater than a selected intensity transition threshold. The method further includes applying crosstalk correction to pixels that are identified as being in at least one of the first set and the second set.Type: GrantFiled: July 18, 2012Date of Patent: July 14, 2015Assignee: QUALCOMM IncorporatedInventors: Gokce Dane, Vasudev Bhaskaran
-
Patent number: 9007395Abstract: Devices, systems, apparatuses, methods, and other embodiments associated with bit resolution enhancement are described. In one embodiment, an apparatus includes logic configured to produce a high-resolution pixel from a low-resolution pixel. The apparatus includes logic configured to classify the high-resolution pixel as being in a smooth region of an image based on at least one of a gradient value and a variance value associated with the low-resolution pixel. The apparatus includes logic configured to selectively re-classify the high-resolution pixel as not being in the smooth region of the image based on a set of neighboring high-resolution pixels associated with high-resolution pixel. The apparatus includes logic configured to selectively filter the high-resolution pixel based on whether the high-resolution pixel remains classified as being in the smooth region of the image.Type: GrantFiled: November 10, 2009Date of Patent: April 14, 2015Assignee: Marvell World Trade Ltd.Inventors: Vasudev Bhaskaran, Nikhil Balram, Sujith Srinivasan, Sanjay Garg