Patents by Inventor Tommi Koivisto
Tommi Koivisto 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: 20240135173Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: ApplicationFiled: June 27, 2023Publication date: April 25, 2024Inventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Patent number: 11790230Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: GrantFiled: April 18, 2022Date of Patent: October 17, 2023Assignee: NVIDIA CorporationInventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Patent number: 11704890Abstract: In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.Type: GrantFiled: November 9, 2021Date of Patent: July 18, 2023Assignee: NVIDIA CorporationInventors: Yilin Yang, Bala Siva Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20220253706Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: ApplicationFiled: April 18, 2022Publication date: August 11, 2022Inventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Patent number: 11308338Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: GrantFiled: December 27, 2019Date of Patent: April 19, 2022Assignee: NVIDIA CorporationInventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20220108465Abstract: In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.Type: ApplicationFiled: November 9, 2021Publication date: April 7, 2022Inventors: Yilin Yang, Bala Siva Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20220101635Abstract: In various examples, detected object data representative of locations of detected objects in a field of view may be determined. One or more clusters of the detected objects may be generated based at least in part on the locations and features of the cluster may be determined for use as inputs to a machine learning model(s). A confidence score, computed by the machine learning model(s) based at least in part on the inputs, may be received, where the confidence score may be representative of a probability that the cluster corresponds to an object depicted at least partially in the field of view. Further examples provide approaches for determining ground truth data for training object detectors, such as for determining coverage values for ground truth objects using associated shapes, and for determining soft coverage values for ground truth objects.Type: ApplicationFiled: November 22, 2021Publication date: March 31, 2022Inventors: Tommi Koivisto, Pekka Janis, Tero Kuosmanen, Timo Roman, Sriya Sarathy, William Zhang, Nizar Assaf, Colin Tracey
-
Patent number: 11210537Abstract: In various examples, detected object data representative of locations of detected objects in a field of view may be determined. One or more clusters of the detected objects may be generated based at least in part on the locations and features of the cluster may be determined for use as inputs to a machine learning model(s). A confidence score, computed by the machine learning model(s) based at least in part on the inputs, may be received, where the confidence score may be representative of a probability that the cluster corresponds to an object depicted at least partially in the field of view. Further examples provide approaches for determining ground truth data for training object detectors, such as for determining coverage values for ground truth objects using associated shapes, and for determining soft coverage values for ground truth objects.Type: GrantFiled: February 15, 2019Date of Patent: December 28, 2021Assignee: NVIDIA CorporationInventors: Tommi Koivisto, Pekka Janis, Tero Kuosmanen, Timo Roman, Sriya Sarathy, William Zhang, Nizar Assaf, Colin Tracey
-
Patent number: 11182916Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: GrantFiled: December 27, 2019Date of Patent: November 23, 2021Assignee: NVIDIA CorporationInventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20210272304Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: ApplicationFiled: December 27, 2019Publication date: September 2, 2021Inventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20200210726Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.Type: ApplicationFiled: December 27, 2019Publication date: July 2, 2020Inventors: Yilin Yang, Bala Siva Sashank Jujjavarapu, Pekka Janis, Zhaoting Ye, Sangmin Oh, Minwoo Park, Daniel Herrera Castro, Tommi Koivisto, David Nister
-
Publication number: 20190258878Abstract: In various examples, detected object data representative of locations of detected objects in a field of view may be determined. One or more clusters of the detected objects may be generated based at least in part on the locations and features of the cluster may be determined for use as inputs to a machine learning model(s). A confidence score, computed by the machine learning model(s) based at least in part on the inputs, may be received, where the confidence score may be representative of a probability that the cluster corresponds to an object depicted at least partially in the field of view. Further examples provide approaches for determining ground truth data for training object detectors, such as for determining coverage values for ground truth objects using associated shapes, and for determining soft coverage values for ground truth objects.Type: ApplicationFiled: February 15, 2019Publication date: August 22, 2019Inventors: Tommi Koivisto, Pekka Janis, Tero Kuosmanen, Timo Roman, Sriya Sarathy, William Zhang, Nizar Assaf, Colin Tracey
-
Publication number: 20190251442Abstract: A neural network includes at least a first network layer that includes a first set of filters and a second network layer that includes a second set of filters. Notably, a filter was removed from the first network layer. A bias associated with a different filter included in the second set of filters compensates for a different bias associated with the filter that was removed from the first network layer.Type: ApplicationFiled: January 11, 2019Publication date: August 15, 2019Inventors: Tommi KOIVISTO, Pekka JÄNIS
-
Patent number: 10284269Abstract: A communications system has a cellular structure including a base station that is located within a cell of the cellular structure and provides an elevation beamforming transmission based on a set of elevation precoding matrix indicator offsets in an elevation codebook. The communications system also includes user equipment that is located within the cell and coupled to the base station to receive the set of elevation precoding matrix indicator offsets and a set of reference signals to provide channel quality and inter-cell interference measurements, wherein a selected channel quality indicator is based on an increase in channel quality with respect to inter-cell interference at the user equipment and corresponds to one of the set of elevation precoding matrix indicator offsets. A method of operating a communications system having a cellular structure is also provided.Type: GrantFiled: January 25, 2016Date of Patent: May 7, 2019Assignee: Nvidia CorporationInventors: Pekka Janis, Tommi Koivisto, Kari Hamalainen
-
Patent number: 10044489Abstract: A method includes operating a network access node to determine, in cooperation with at least one other network access node, a coordinated scheduling granularity having a plurality of physical resource blocks; and signaling an indication of the determined coordinated scheduling granularity to at least one mobile device served by the network access node for use in enhancing estimation at a receiver of the mobile device, such as when estimating an interference covariance matrix. Apparatus for performing the method is also disclosed, as are mobile device methods and apparatus for receiving and using the signaling.Type: GrantFiled: October 21, 2011Date of Patent: August 7, 2018Assignee: Nokia Solutions and Networks OyInventors: Kari Pekka Pajukoski, Esa Tapani Tiirola, Kari Juhani Hooli, Tommi Koivisto, Marko Lampinen, Lars Lindh
-
Patent number: 9887751Abstract: A communications system has a cellular structure and the communications system includes a base station that is located within a cell of the cellular structure and employs a Kronecker product of azimuth and elevation precoding vectors for beamforming. Additionally, the communications system includes user equipment that is located within the cell and coupled to the base station to receive a reference channel state information process employing a reference precoding vector for use in a non-reference channel state information process to derive a compensated channel quality indication. A method of operating a communications system is also included.Type: GrantFiled: January 25, 2016Date of Patent: February 6, 2018Assignee: Nvidia CorporationInventors: Pekka Janis, Tommi Koivisto, Kari Hamalainen
-
Patent number: 9867148Abstract: A method including receiving at a user equipment first power information for transmitting to a first base station, receiving at the user equipment second power information for transmitting to a second base station, causing said user equipment to transmit to said first base station with a first power less than or equal to a first maximum power dependent on said first power information and causing said user equipment to transmit to said second base station with a second power less than or equal to a second maximum power dependent on said second power information, such that said first and second power does not exceed a total power allowed for said user equipment.Type: GrantFiled: January 31, 2014Date of Patent: January 9, 2018Assignee: Nokia Solutions and Networks OyInventors: Jari Olavi Lindholm, Kari Juhani Hooli, Timo Erkki Lunttila, Esa Tapani Tiirola, Tommi Koivisto
-
Patent number: 9853782Abstract: Information about a set of cells potentially taking part in coordinated multipoint transmission is obtained, information about at least one of reference signal configuration and data region size configuration for each of the set of cells is obtained, further an indication of which of the reference signals configuration and the data region size configuration should be assumed in resource mapping for data is obtained, and data according to the resource mapping is received. Information about at least one of reference signal configuration and data region size configuration for each cell of a set of cells potentially taking part in coordinated multipoint transmission is provided, and an indication of which of the reference signals configuration and the data region size configuration should be assumed in resource mapping for data is provided.Type: GrantFiled: February 10, 2016Date of Patent: December 26, 2017Assignee: Avago Technologies General IP (Singapore) Pte. Ltd.Inventor: Tommi Koivisto
-
Patent number: 9831999Abstract: Methods for operating a small cell in a discontinued reception (DRX) mode include maintaining the small cell in a discontinuous transmission (DTX) mode during a first time period having a plurality of first time slots. The methods include transmitting common reference signals in a predetermined number of second time slots prior to the first time slots and in a predetermined number of third time slots following commencement of the first time slots. The methods include discontinuing transmission of the common reference signals and common channel signals if mobile devices are in a discontinuous reception mode during the first time period. The methods include discontinuing transmission of the common reference signals during a predetermined number of fourth time slots following commencement of the first time period if there is no dedicated common transmission to the mobile devices.Type: GrantFiled: May 8, 2015Date of Patent: November 28, 2017Assignee: Nvidia CorporationInventors: Tommi Koivisto, Tero Kuosmanen, Timo Roman
-
Publication number: 20170195978Abstract: A method comprising including receiving at a user equipment first power information for transmitting to a first base station, receiving at the user equipment second power information for transmitting to a second base station, causing said user equipment to transmit to said first base station with a first power less than or equal to a first maximum power dependent on said first power information and causing said user equipment to transmit to said second base station with a second power less than or equal to a second maximum power dependent on said second power information, such that said first and second power does not exceed a total power allowed for said user equipmentType: ApplicationFiled: January 31, 2014Publication date: July 6, 2017Inventors: Jari Olavi Lindholm, Kari Juhani HOOLI, Timo Erkki LUNTTILA, Esa Tapani TIIROLA, Tommi KOIVISTO