Patents by Inventor Vijay Mahadevan

Vijay Mahadevan 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: 10936920
    Abstract: A system trains and applies a machine learning model to label maps of a region. Various data modalities are combined as inputs for multiple data tiles used to characterize a region for a geographical map. Each data modality reflects sensor data captured in different ways. Some data modalities include aerial imagery, point cloud data, and location trace data. The different data modalities are captured independently and then aggregated using machine learning models to determine map labeling information about tiles in the region. Data is ingested by the system and corresponding tiles are identified. A tile is represented by a feature vector of different data types related to the various data modalities, and values from the ingested data are added to the feature vector for the tile. Models can be trained to predict characteristics of a region using these various types of input.
    Type: Grant
    Filed: June 17, 2019
    Date of Patent: March 2, 2021
    Assignee: Uber Technologies, Inc.
    Inventors: Timo Pekka Pylvaenaeinen, Aditya Sarawgi, Vijay Mahadevan, Vasudev Parameswaran, Mohammed Waleed Kadous
  • Patent number: 10896355
    Abstract: Disclosed are systems and methods for automatic selection of canonical digital images from a large corpus of digital images, such as the corpus of digital images available on the web, for an entity, such as and without limitation a person, a point of interest, object, etc. The automated, unsupervised approach for selecting a diverse set of high quality, canonical digital images, is well suited for processing a large corpus of digital images. A set of canonical digital images identified for an entity can be retrieved in response to a digital image request for digital images depicting the entity.
    Type: Grant
    Filed: December 21, 2018
    Date of Patent: January 19, 2021
    Assignee: VERIZON MEDIA INC.
    Inventors: Sachin Sudhakar Farfade, Vijay Mahadevan, Ayman Kaheel, Ayyappan Arasu, Venkat Kumar Reddy Barakam, Jay Kiran Mahadeokar
  • Publication number: 20200403960
    Abstract: A method and apparatus of a device that starts an address resolution service on a network element after a boot-up of this network element is described. In an exemplary embodiment, the network element sends an indication of the boot-up. The network element further sends a request for an address resolution table and receives a reply with the requested address resolution table. In addition, the network element starts the address resolution service using the requested address resolution table.
    Type: Application
    Filed: June 10, 2020
    Publication date: December 24, 2020
    Inventors: Vijay Mahadevan, Dileep Honsoge Ramesh, Adam James Sweeney
  • Patent number: 10839245
    Abstract: A structured document analyzer that associates keys and values in structured documents based on key, value, and key-value container bounding boxes. A trained machine learning model analyzes images of structured documents to determine bounding boxes for keys, values, and key-value containers in the images with confidence scores for the classifications. For each image, duplicate bounding boxes are removed, and then a set of key-value containers are selected and sorted based on the confidence scores. For each key-value container, a best key and value are determined for the container based on overlap of the key and value bounding boxes with the container bounding box and the confidence scores. Optical character recognition may be performed on the image to determine text for the keys and values.
    Type: Grant
    Filed: March 25, 2019
    Date of Patent: November 17, 2020
    Assignee: Amazon Technologies, Inc.
    Inventors: Guneet Singh Dhillon, Vijay Mahadevan, Yuting Zhang, Meng Wang, Gangadhar Payyavula, Viet Cuong Nguyen, Rahul Bhotika, Stefano Soatto
  • Patent number: 10762644
    Abstract: Techniques for multiple object tracking in video are described in which the outputs of neural networks are combined within a Bayesian framework. A motion model is applied to a probability distribution representing the estimated current state of a target object being tracked to predict the state of the target object in the next frame. A state of an object can include one or more features, such as the location of the object in the frame, a velocity and/or acceleration of the object across frames, a classification of the object, etc. The prediction of the state of the target object in the next frame is adjusted by a score based on the combined outputs of neural networks that process the next frame.
    Type: Grant
    Filed: December 13, 2018
    Date of Patent: September 1, 2020
    Assignee: Amazon Technologies, Inc.
    Inventors: Vijay Mahadevan, Stefano Soatto
  • Patent number: 10740619
    Abstract: A media item comprising a set of frames is received by a feature extraction system. A frame predictor is executed on each frame of the set of frames. An error representation is extracted for each frame of the set of frames during the execution of the frame predictor. An error-based feature vector is generated from the error representations associated with each frame of the set of frames. A seed media item is identified having a first error-based feature vector. A similarity score is determined among the first error-based feature vector and each error-based feature vector of a set of error-based feature vectors. A subset of error-based feature vectors, hence a subset of corresponding media items, is selected based on similarity score.
    Type: Grant
    Filed: August 31, 2018
    Date of Patent: August 11, 2020
    Assignee: Uber Technologies, Inc.
    Inventors: Pegah Massoudifar, Vijay Mahadevan, Jonathan Berliner
  • Patent number: 10721206
    Abstract: A method and apparatus of a device that starts an address resolution service on a network element after a boot-up of this network element is described. In an exemplary embodiment, the network element sends an indication of the boot-up. The network element further sends a request for an address resolution table and receives a reply with the requested address resolution table. In addition, the network element starts the address resolution service using the requested address resolution table.
    Type: Grant
    Filed: February 2, 2016
    Date of Patent: July 21, 2020
    Assignee: Arista Networks, Inc.
    Inventors: Vijay Mahadevan, Dileep Honsoge Ramesh, Adam James Sweeney
  • Publication number: 20190385010
    Abstract: A system trains and applies a machine learning model to label maps of a region. Various data modalities are combined as inputs for multiple data tiles used to characterize a region for a geographical map. Each data modality reflects sensor data captured in different ways. Some data modalities include aerial imagery, point cloud data, and location trace data. The different data modalities are captured independently and then aggregated using machine learning models to determine map labeling information about tiles in the region. Data is ingested by the system and corresponding tiles are identified. A tile is represented by a feature vector of different data types related to the various data modalities, and values from the ingested data are added to the feature vector for the tile. Models can be trained to predict characteristics of a region using these various types of input.
    Type: Application
    Filed: June 17, 2019
    Publication date: December 19, 2019
    Inventors: Timo Pekka Pylvaenaeinen, Aditya Sarawgi, Vijay Mahadevan, Vasudev Parameswaran, Mohammed Waleed Kadous
  • Publication number: 20190156125
    Abstract: A media item comprising a set of frames is received by a feature extraction system. A frame predictor is executed on each frame of the set of frames. An error representation is extracted for each frame of the set of frames during the execution of the frame predictor. An error-based feature vector is generated from the error representations associated with each frame of the set of frames. A seed media item is identified having a first error-based feature vector. A similarity score is determined among the first error-based feature vector and each error-based feature vector of a set of error-based feature vectors. A subset of error-based feature vectors, hence a subset of corresponding media items, is selected based on similarity score.
    Type: Application
    Filed: August 31, 2018
    Publication date: May 23, 2019
    Inventors: Pegah Massoudifar, Vijay Mahadevan, Jonathan Berliner
  • Patent number: 10163041
    Abstract: Disclosed are systems and methods for automatic selection of canonical digital images from a large corpus of digital images, such as the corpus of digital images available on the web, for an entity, such as and without limitation a person, a point of interest, object, etc. The automated, unsupervised approach for selecting a diverse set of high quality, canonical digital images, is well suited for processing a large corpus of digital images. A set of canonical digital images identified for an entity can be retrieved in response to a digital image request for digital images depicting the entity.
    Type: Grant
    Filed: June 30, 2016
    Date of Patent: December 25, 2018
    Assignee: OATH INC.
    Inventors: Sachin Sudhakar Farfade, Vijay Mahadevan, Ayman Kaheel, Ayyappan Arasu, Venkat Kumar Reddy Barakam, Jan Kiran Mahadeokar
  • Publication number: 20180260417
    Abstract: Disclosed are systems and methods for improving interactions with and between computers in content searching, generating, hosting and/or providing systems supported by or configured with personal computing devices, servers and/or platforms. The systems interact to identify and retrieve data within or across platforms, which can be used to improve the quality of data used in processing interactions between or among processors in such systems. The disclosure provides a novel, computerized framework for automatically selecting the most definitive, precise and high-quality content files corresponding to POIs. The disclosed systems and methods utilize the performance of visual comparisons with a set of definitive content files of a given POI, and by incorporating visual aesthetic features as a factor of such comparisons, a search result is identified that down-weights imprecise and poor quality content files of a given POI, and ensures that only high quality, accurate content files are selected or identified.
    Type: Application
    Filed: March 7, 2017
    Publication date: September 13, 2018
    Inventors: Vijay MAHADEVAN, Sachin Sudhakar FARFADE, Jay Kiran MAHADEOKAR, Ayyappan ARASU, Venkat Kumar Reddy BARAKAM, Ayman KAHEEL
  • Publication number: 20180005088
    Abstract: Disclosed are systems and methods for automatic selection of canonical digital images from a large corpus of digital images, such as the corpus of digital images available on the web, for an entity, such as and without limitation a person, a point of interest, object, etc. The automated, unsupervised approach for selecting a diverse set of high quality, canonical digital images, is well suited for processing a large corpus of digital images. A set of canonical digital images identified for an entity can be retrieved in response to a digital image request for digital images depicting the entity.
    Type: Application
    Filed: June 30, 2016
    Publication date: January 4, 2018
    Inventors: Sachin Sudhakar Farfade, Vijay Mahadevan, Ayman Kaheel, Ayyappan Arasu, Venkat Kumar Reddy Barakam, Jan Kiran Mahadeokar
  • Publication number: 20160255043
    Abstract: A method and apparatus of a device that starts an address resolution service on a network element after a boot-up of this network element is described. In an exemplary embodiment, the network element sends an indication of the boot-up. The network element further sends a request for an address resolution table and receives a reply with the requested address resolution table. In addition, the network element starts the address resolution service using the requested address resolution table.
    Type: Application
    Filed: February 2, 2016
    Publication date: September 1, 2016
    Inventors: Vijay MAHADEVAN, Dileep Honsoge RAMESH, Adam James Sweeney
  • Patent number: 9210327
    Abstract: Users are provided with feedback regarding blurriness of an image in real-time. When an image is received, a blur score is automatically generated in addition to a visual that indicates the extent of blurriness across the picture. The blur score is calculated by aggregating an image_blur_score and optionally a motion_blur_score. A user can also be provided with suggestions on improving image sharpness and help in determining if another image needs to be taken.
    Type: Grant
    Filed: December 2, 2013
    Date of Patent: December 8, 2015
    Assignee: YAHOO! INC.
    Inventors: Gaurav Aggarwal, Nikhil Rasiwasia, Kshitiz Garg, Vijay Mahadevan
  • Publication number: 20150156419
    Abstract: Users are provided with feedback regarding blurriness of an image in real-time. When an image is received, a blur score is automatically generated in addition to a visual that indicates the extent of blurriness across the picture. The blur score is calculated by aggregating an image_blur_score and optionally a motion_blur_score. A user can also be provided with suggestions on improving image sharpness and help in determining if another image needs to be taken.
    Type: Application
    Filed: December 2, 2013
    Publication date: June 4, 2015
    Applicant: YAHOO! INC.
    Inventors: Gaurav Aggarwal, Nikhil Rasiwasia, Kshitiz Garg, Vijay Mahadevan
  • Patent number: 9025673
    Abstract: The disclosure is directed to techniques for evaluating temporal quality of encoded video. Instead of estimating jerkiness based solely on frame rate or motion activity, the number of consecutive dropped frames forms a basic estimation unit. Several human visual system factors, such as sensitivity to temporal quality fluctuation and motion activity, have been taken into account to make the predicted jerkiness more consistent with the actual human visual response. The temporal quality metric can be used to estimate human perceived discomfort that is introduced by temporal discontinuity under various combinations of video shots, motion activity and local quality fluctuations. The techniques can be applied in two modes: (1) bitstream or (2) pixel mode. The quality metric can be used to evaluate temporal quality, or to control encoding or decoding characteristics to enhance temporal quality.
    Type: Grant
    Filed: June 8, 2006
    Date of Patent: May 5, 2015
    Assignee: QUALCOMM Incorporated
    Inventors: Kai-Chieh Yang, Khaled Helmi El-Maleh, Vijay Mahadevan
  • Patent number: 8744203
    Abstract: The disclosure is directed to decoder-side region-of-interest (ROI) video processing. A video decoder determines whether ROI assistance information is available. If not, the decoder defaults to decoder-side ROI processing. The decoder-side ROI processing may estimate the reliability of ROI extraction in the bitstream domain. If ROI reliability is favorable, the decoder applies bitstream domain ROI extraction. If ROI reliability is unfavorable, the decoder applies pixel domain ROI extraction. The decoder may apply different ROI extraction processes for intra-coded (I) and inter-coded (P or B) data. The decoder may use color-based ROI generation for intra-coded data, and coded block pattern (CBP)-based ROI generation for inter-coded data. ROI refinement may involve shape-based refinement for intra-coded data, and motion- and color-based refinement for inter-coded data.
    Type: Grant
    Filed: April 27, 2012
    Date of Patent: June 3, 2014
    Assignee: QUALCOMM Incorporated
    Inventors: Khaled Helmi El-Maleh, Vijay Mahadevan, Haohong Wang
  • Patent number: 8582660
    Abstract: This disclosure is directed to techniques for selective video frame rate upconversion (FRUC) in a video decoder. A video decoder selectively enables or disables FRUC based on one or more adaptive criteria. The adaptive criteria may be selected to indicate whether FRUC is likely to introduce spatial artifacts. Adaptive criteria may include a motion activity threshold, a mode decision threshold, or both. The criteria are adaptive, rather than fixed. When the criteria indicate that a frame includes excessive motion or new content, the decoder disables FRUC.
    Type: Grant
    Filed: August 29, 2006
    Date of Patent: November 12, 2013
    Assignee: QUALCOMM Incorporated
    Inventors: Vijay Mahadevan, Brijesh Pillai, Khaled Helmi El-Maleh, Fang Shi, Vijayalakshmi R. Raveendran
  • Patent number: 8315466
    Abstract: The disclosure is directed to decoder-side region-of-interest (ROI) video processing. A video decoder determines whether ROI assistance information is available. If not, the decoder defaults to decoder-side ROI processing. The decoder-side ROI processing may estimate the reliability of ROI extraction in the bitstream domain. If ROI reliability is favorable, the decoder applies bitstream domain ROI extraction. If ROI reliability is unfavorable, the decoder applies pixel domain ROI extraction. The decoder may apply different ROI extraction processes for intra-coded (I) and inter-coded (P or B) data. The decoder may use color-based ROI generation for intra-coded data, and coded block pattern (CBP)-based ROI generation for inter-coded data. ROI refinement may involve shape-based refinement for intra-coded data, and motion- and color-based refinement for inter-coded data.
    Type: Grant
    Filed: December 22, 2006
    Date of Patent: November 20, 2012
    Assignee: QUALCOMM Incorporated
    Inventors: Khaled Helmi El-Maleh, Vijay Mahadevan, Haohong Wang
  • Publication number: 20120213409
    Abstract: The disclosure is directed to decoder-side region-of-interest (ROI) video processing. A video decoder determines whether ROI assistance information is available. If not, the decoder defaults to decoder-side ROI processing. The decoder-side ROI processing may estimate the reliability of ROI extraction in the bitstream domain. If ROI reliability is favorable, the decoder applies bitstream domain ROI extraction. If ROI reliability is unfavorable, the decoder applies pixel domain ROI extraction. The decoder may apply different ROI extraction processes for intra-coded (I) and inter-coded (P or B) data. The decoder may use color-based ROI generation for intra-coded data, and coded block pattern (CBP)-based ROI generation for inter-coded data. ROI refinement may involve shape-based refinement for intra-coded data, and motion- and color-based refinement for inter-coded data.
    Type: Application
    Filed: April 27, 2012
    Publication date: August 23, 2012
    Applicant: QUALCOMM Incorporated
    Inventors: Khaled Helmi El-Maleh, Vijay Mahadevan, Haohong Wang