Patents by Inventor Harri Valpola
Harri Valpola 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: 11720795Abstract: Disclosed is a neural network structure enabling efficient training of the network and a method thereto. The structure is a ladder-type structure wherein one or more lateral input(s) is/are taken to decoding functions. By minimizing one or more cost function(s) belonging to the structure the neural network structure may be trained in an efficient way.Type: GrantFiled: November 26, 2014Date of Patent: August 8, 2023Assignee: Canary Capital LLCInventor: Harri Valpola
-
Patent number: 11669056Abstract: The invention relates to a method for generating a control system for a target system, wherein: operational data is received; a first neural model component is trained with the received operational data for generating a prediction on a state of the target system based on the received operational data; a second neural model component is trained with the operational data for generating a regularizer for use in inverting the first neural model component; and the control system is generated by inverting the first neural model component by optimization and arranging to apply the regularizer generated with the second neural model component in the optimization. The invention relates also to a system and a computer program product.Type: GrantFiled: October 31, 2018Date of Patent: June 6, 2023Assignee: CANARY CAPITAL LLCInventors: Harri Valpola, Eva Koppali
-
Patent number: 11620511Abstract: Disclosed is a computer-implemented method for training a neural network system including an original neural network and a label generator. The method is based on an idea that the neural network system is trained by a sequence of training steps where at each training step at least one of a plurality of operations is performed and each of the operations gets performed at least once during training of the neural network system. Also disclosed are a neural network system and a computer program product.Type: GrantFiled: February 14, 2018Date of Patent: April 4, 2023Assignee: Canary Capital LLCInventor: Harri Valpola
-
Patent number: 11568208Abstract: Disclosed is a computer-implemented method for estimating an uncertainty of a prediction generated by a machine learning system, the method including: receiving first data; training a first machine learning model component of a machine learning system with the received first data, the first machine learning model component is trained to generate a prediction; generating an uncertainty estimate of the prediction; training a second machine learning model component of the machine learning system with second data, the second machine learning model component is trained to generate a calibrated uncertainty estimate of the prediction. Also disclosed is a corresponding system.Type: GrantFiled: November 8, 2019Date of Patent: January 31, 2023Assignee: Canary Capital LLCInventor: Harri Valpola
-
Patent number: 11481585Abstract: Disclosed is a computer-implemented method for segmenting input data. In the method a plurality of tags is generated; the input data is masked with the plurality of tags; a plurality of output reconstructions is generated by inputting the plurality of masked input data to one of the following: a denoising neural network, a variational autoencoder; a plurality of values representing distances of each plurality of output reconstructions to the input data are determined; a plurality of updated versions of input data is generated by applying at least one of the determined values representing distances of each plurality of output reconstructions to the input data; and updated output reconstructions are generated by inputting the plurality of updated versions of input data to one of the networks. Also disclosed is a method for training the network and a processing unit.Type: GrantFiled: May 19, 2017Date of Patent: October 25, 2022Assignee: Canary Capital LLCInventors: Harri Valpola, Klaus Greff
-
Publication number: 20210341884Abstract: The invention relates to a method for generating a control system for a target system, wherein: operational data is received; a first neural model component is trained with the received operational data for generating a prediction on a state of the target system based on the received operational data; a second neural model component is trained with the operational data for generating a regularizer for use in inverting the first neural model component; and the control system is generated by inverting the first neural model component by optimization and arranging to apply the regularizer generated with the second neural model component in the optimization. The invention relates also to a system and a computer program product.Type: ApplicationFiled: October 31, 2018Publication date: November 4, 2021Inventors: Harri VALPOLA, Eva KOPPALI
-
Patent number: 11138468Abstract: A method for generating an output signal of a system based on input data received by the system includes receiving training data and training a neural network for generating the output signal by optimizing a primary cost function and an auxiliary cost function and modulating the auxiliary cost function with a gradient-based attention mask during the training.Type: GrantFiled: May 21, 2018Date of Patent: October 5, 2021Assignee: Canary Capital LLCInventors: Matti Herranen, Harri Valpola
-
Publication number: 20200234116Abstract: Disclosed is a computer-implemented method for training a neural network system including an original neural network and a label generator. The method is based on an idea that the neural network system is trained by a sequence of training steps where at each training step at least one of a plurality of operations is performed and each of the operations gets performed at least once during training of the neural network system. Also disclosed are a neural network system and a computer program product.Type: ApplicationFiled: February 14, 2018Publication date: July 23, 2020Inventor: Harri VALPOLA
-
Publication number: 20200202172Abstract: Disclosed is a method for generating an output signal of a system based on input data received by the system, the method including: receiving training data; training a neural network for generating the output signal by optimizing a primary cost function and an auxiliary cost function and modulating the auxiliary cost function with a gradient-based attention mask during the training; wherein the method further including: receiving the input data; inputting the received input data to the trained neural network: generating the output signal of the system in accordance with a processing of the received input data with the trained neural network. Also disclosed is a system and a computer program product.Type: ApplicationFiled: May 21, 2018Publication date: June 25, 2020Inventors: Matti HERRANEN, Harri VALPOLA
-
Publication number: 20200150601Abstract: Disclosed is a method for controlling a target system, the method including: receiving data of at least one source system, training a first machine learning model component with the received data to generate a prediction on a state of the target system, generating an uncertainty estimate of the prediction, training a second machine learning model machine learning component with the received data to generate a calibrated uncertainty estimate of the prediction; and the method further including: receiving an operational data of the target system, controlling the target system by way of selecting a control action by optimization using the first machine learning model component and arranging to apply the calibrated uncertainty estimate generated with the second machine learning model component in the optimization.Type: ApplicationFiled: November 8, 2019Publication date: May 14, 2020Inventor: Harri VALPOLA
-
Publication number: 20200151547Abstract: Disclosed is a computer-implemented method for estimating an uncertainty of a prediction generated by a machine learning system, the method including: receiving first data; training a first machine learning model component of a machine learning system with the received first data, the first machine learning model component is trained to generate a prediction; generating an uncertainty estimate of the prediction; training a second machine learning model component of the machine learning system with second data, the second machine learning model component is trained to generate a calibrated uncertainty estimate of the prediction. Also disclosed is a corresponding system.Type: ApplicationFiled: November 8, 2019Publication date: May 14, 2020Inventor: Harri VALPOLA
-
Publication number: 20190220691Abstract: Disclosed is a computer-implemented method for segmenting input data. In the method a plurality of tags is generated; the input data is masked with the plurality of tags; a plurality of output reconstructions is generated by inputting the plurality of masked input data to one of the following: a denoising neural network, a variational autoencoder; a plurality of values representing distances of each plurality of output reconstructions to the input data are determined; a plurality of updated versions of input data is generated by applying at least one of the determined values representing distances of each plurality of output reconstructions to the input data; and updated output reconstructions are generated by inputting the plurality of updated versions of input data to one of the networks. Also disclosed is a method for training the network and a processing unit.Type: ApplicationFiled: May 19, 2017Publication date: July 18, 2019Applicant: Curious AI OyInventors: Harri VALPOLA, Klaus GREFF
-
Publication number: 20170330076Abstract: Disclosed is a neural network structure enabling efficient training of the network and a method thereto. The structure is a ladder-type structure wherein one or more lateral input(s) is/are taken to decoding functions. By minimizing one or more cost function(s) belonging to the structure the neural network structure may be trained in an efficient way.Type: ApplicationFiled: November 26, 2014Publication date: November 16, 2017Inventor: Harri VALPOLA
-
Patent number: 9050719Abstract: A method in which an apparatus receives first sensor data from first sensors and determines a target position from the data, the target position may be a position in space or an orientation of a gripper in a robot arm First instructions are issued to the robot arm or the gripper in order to move a gripper to the target position. Force feedback sensor data is received from force feedback sensors associated with either the robot arm or the gripper or from the first sensors. A failure in carrying out the first instructions is determined. Second sensor data is received from the at least one first sensor. Successful gripping of an object is determined from the second sensor data.Type: GrantFiled: May 5, 2011Date of Patent: June 9, 2015Assignee: ZENROBOTICS OYInventors: Harri Valpola, Tuomas Lukka
-
Publication number: 20140088765Abstract: The invention relates to a method and system for invalidating sensor measurements after a sorting action on a target area of a robot sorting system. In the method there are obtained sensor measurements using sensors from a target area. A first image is captured of the target area using a sensor over the target area. A first sorting action is performed in the target area using a robot arm based on the sensor measurements and the first image. Thereupon, a second image of the target area is captured using a sensor over the target area. The first and the second images are compared to determine invalid areas in the target area. The invalid areas are avoided in future sorting actions based on the sensor measurements.Type: ApplicationFiled: March 28, 2012Publication date: March 27, 2014Applicant: ZENROBOTICS OYInventors: Harri Valpola, Tuomas Lukka
-
Publication number: 20130266205Abstract: The invention relates to a method and system for recognizing physical objects. In the method an object is gripped with a gripper, which is attached to a robot arm or mounted separately. Using an image sensor, a plurality of source images of an area comprising the object is captured while the object is moved with the robot arm. The camera is configured to move along the gripper, attached to the gripper or otherwise able to monitor the movement of the gripper. Moving image elements are extracted from the plurality of source images by computing a variance image from the source images and forming a filtering image from the variance image. A result image is obtained by using the filtering image as a bitmask. The result image is used for classifying the gripped object.Type: ApplicationFiled: October 12, 2011Publication date: October 10, 2013Applicant: ZenRobotics OyInventor: Harri Valpola
-
Publication number: 20130151007Abstract: A method in which an apparatus receives first sensor data from first sensors and determines a target position from the data, the target position may be a position in space or an orientation of a gripper in a robot arm First instructions are issued to the robot arm or the gripper in order to move a gripper to the target position. Force feedback sensor data is received from force feedback sensors associated with either the robot arm or the gripper or from the first sensors. A failure in carrying out the first instructions is determined Second sensor data is received from the at least one first sensor. Successful gripping of an object is determined from the second sensor data.Type: ApplicationFiled: May 5, 2011Publication date: June 13, 2013Applicant: ZENROBOTICS OYInventors: Harri Valpola, Tuomas Lukka
-
Patent number: 8332336Abstract: In the method of the invention, information is selected in a processing unit. The end result to be achieved is final feature information to be used for performing a task. In the method, input information to the processing unit is treated with the intention to find useful information from the input information in view of the task. Feature information representing characteristics of the input information is formed from the input data, the values of the feature information being further processed on the basis of their utility. The method is mainly characterized in that a preliminary prediction is formed from a first context of a sot of input information by means of first parameters and in that the first primary input of the set of input data is preprocessed by forming calculation results by means of second parameters.Type: GrantFiled: February 22, 2008Date of Patent: December 11, 2012Assignee: Zenrobotics OyInventors: Harri Valpola, Antti Yli-Krekola
-
Publication number: 20100088259Abstract: In the method of the invention, information is selected in a processing unit. The end result to be achieved is final feature information to be used for performing a task. In the method, input information to the processing unit is treated with the intention to find useful information from the input information in view of the task. The input information consists of a primary input and a context representing general information related to the task to be performed. Feature information representing characteristics of the input information is formed from the input data, the values of the feature information being further processed on the basis of their utility. The method is mainly characterized in that a preliminary prediction is formed from a first context of a set of input information by means of first parameters and in that the first primary input of the set of input data is preprocessed by forming calculation results by means of second parameters.Type: ApplicationFiled: February 22, 2008Publication date: April 8, 2010Applicant: ZENROBOTICS OYInventors: Harri Valpola, Antti Yli-Krekola