-
公开(公告)号：US20240149440A1
公开(公告)日：2024-05-09
申请号：US18416283
申请日：2024-01-18
Applicant: Embodied Intelligence Inc.
Inventor： Yan Duan , Haoran Tang , Yide Shentu , Nikhil Mishra , Xi Chen
CPC classification number: B25J9/161 , B25J9/1612 , B25J9/163 , B25J9/1653 , B25J15/0658
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to an artificial neural network training method that does not require extensive training data or time expenditure. The few-shot training model disclosed herein includes attempting to pick up items and, in response to a failed pick up attempt, transferring and generalizing information to similar regions to improve probability of success in future attempts. In some implementations, the training method is used to robotic device for picking items from a bin and perturbing items in a bin. When no picking strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points. In some implementations, the device may include one or more computer-vision systems.
-
公开(公告)号：US20210233258A1
公开(公告)日：2021-07-29
申请号：US17161399
申请日：2021-01-28
Applicant: Embodied Intelligence Inc.
Inventor： Maximilian Sieb , Nikhil Mishra , Rocky Duan
Abstract: Various embodiments of the present technology generally relate to robotic devices, computer vision, and artificial intelligence. More specifically, some embodiments relate to object tracking using neural networks and computer vision systems. In some embodiments, a computer vision system for object tracking captures one or more images of a first scene, wherein the first scene corresponds to a first location, identifies a distinct object in the first scene based on the one or more first images, directs a robotic device to move the distinct object from the first location to a second location, captures one or more second images of a second scene, wherein the second scene corresponds to the second location, and determines if the distinct objects is in the second scene based on the one or more second images.
-
公开(公告)号：US20210069904A1
公开(公告)日：2021-03-11
申请号：US17014545
申请日：2020-09-08
Applicant: Embodied Intelligence, Inc.
Inventor： Yan Duan , Xi Chen , Mostafa Rohaninejad , Nikhil Mishra , Yu Xuan Liu , Andrew Amir Vaziri , Haoran Tang , Yide Shentu , Ian Rust , Carlos Florensa
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to a robotic device for picking items from a bin and perturbing items in a bin. In some implementations, the device may include one or more computer-vision systems. A computer-vision system, in accordance with the present technology, may use at least two two-dimensional images to generate three-dimensional (3D) information about the bin and items in the bin. Based on the 3D information, a strategy for picking up items from the bin is determined. When no strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points and re-attempt to pick up an item.
-
公开(公告)号：US20240342909A1
公开(公告)日：2024-10-17
申请号：US18751576
申请日：2024-06-24
Applicant: Embodied Intelligence Inc.
Inventor： Haoran Tang , Xi Chen , Yan Duan , Nikhil Mishra , Shiyao Wu , Maximilian Sieb , Yide Shentu
CPC classification number: B25J9/1666 , B25J9/1605 , B25J9/163 , B65G61/00 , B25J5/007 , B25J9/1697
Abstract: Various embodiments of the technology described herein generally relate to systems and methods for trajectory optimization with machine learning techniques. More specifically, certain embodiments relate to using neural networks to quickly predict optimized robotic arm trajectories in a variety of scenarios. Systems and methods described herein use deep neural networks to quickly predict optimized robotic arm trajectories according to certain constraints. Optimization, in accordance with some embodiments of the present technology, may include optimizing trajectory geometry and dynamics while satisfying a number of constraints, including staying collision-free and minimizing the time it takes to complete the task.
-
公开(公告)号：US11911901B2
公开(公告)日：2024-02-27
申请号：US17014558
申请日：2020-09-08
Applicant: Embodied Intelligence, Inc.
Inventor： Yan Duan , Haoran Tang , Yide Shentu , Nikhil Mishra , Xi Chen
CPC classification number: B25J9/161 , B25J9/163 , B25J9/1612 , B25J9/1653 , B25J15/0658
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to an artificial neural network training method that does not require extensive training data or time expenditure. The few-shot training model disclosed herein includes attempting to pick up items and, in response to a failed pick up attempt, transferring and generalizing information to similar regions to improve probability of success in future attempts. In some implementations, the training method is used to robotic device for picking items from a bin and perturbing items in a bin. When no picking strategies with high probability of success exist, the robotic device may perturb the contents of the bin to create new available pick-up points. In some implementations, the device may include one or more Computer-vision systems.
-
Patent Agency Ranking
公开(公告)号：US20210233246A1
公开(公告)日：2021-07-29
申请号：US17161344
申请日：2021-01-28
Applicant: Embodied Intelligence Inc.
Inventor： YuXuan Liu , Xi Chen , Nikhil Mishra
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to modeling uncertainty in neural network segmentation predictions of imaged scenes having a plurality of objects. In some embodiments, a computer vision system for guiding robotic picking utilizes a method for uncertainty modeling that comprises receiving one or more images of a scene comprising a plurality of distinct objects, generating a plurality of segmentation predictions each comprising one or more object masks, identifying a predefined confidence requirement, wherein the confidence requirement identifies a minimum amount of required agreement for a region, and outputting one or more object masks based on the confidence requirement. The systems and methods disclosed herein leverage the use of a plurality of hypotheses to create a distribution of possible segmentation outcomes in order model uncertainty associated with image segmentation.
-
公开(公告)号：US12183011B2
公开(公告)日：2024-12-31
申请号：US17161344
申请日：2021-01-28
Applicant: Embodied Intelligence Inc.
Inventor： YuXuan Liu , Xi Chen , Nikhil Mishra
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to modeling uncertainty in neural network segmentation predictions of imaged scenes having a plurality of objects. In some embodiments, a computer vision system for guiding robotic picking utilizes a method for uncertainty modeling that comprises receiving one or more images of a scene comprising a plurality of distinct objects, generating a plurality of segmentation predictions each comprising one or more object masks, identifying a predefined confidence requirement, wherein the confidence requirement identifies a minimum amount of required agreement for a region, and outputting one or more object masks based on the confidence requirement. The systems and methods disclosed herein leverage the use of a plurality of hypotheses to create a distribution of possible segmentation outcomes in order model uncertainty associated with image segmentation.
-
公开(公告)号：US12179363B2
公开(公告)日：2024-12-31
申请号：US17193820
申请日：2021-03-05
Applicant: Embodied Intelligence Inc.
Inventor： Haoran Tang , Xi Chen , Yan Duan , Nikhil Mishra , Shiyao Wu , Maximilian Sieb , Yide Shentu
IPC: B25J9/16
Abstract: Various embodiments of the technology described herein generally relate to systems and methods for trajectory optimization with machine learning techniques. More specifically, certain embodiments relate to using neural networks to quickly predict optimized robotic arm trajectories in a variety of scenarios. Systems and methods described herein use deep neural networks to quickly predict optimized robotic arm trajectories according to certain constraints. Optimization, in accordance with some embodiments of the present technology, may include optimizing trajectory geometry and dynamics while satisfying a number of constraints, including staying collision-free, and minimizing the time it takes to complete the task.
-
公开(公告)号：US20210276187A1
公开(公告)日：2021-09-09
申请号：US17193820
申请日：2021-03-05
Applicant: Embodied Intelligence Inc.
Inventor： Haoran Tang , Xi Chen , Yan Duan , Nikhil Mishra , Shiyao Wu , Maximilian Sieb , Yide Shentu
Abstract: Various embodiments of the technology described herein generally relate to systems and methods for trajectory optimization with machine learning techniques. More specifically, certain embodiments relate to using neural networks to quickly predict optimized robotic arm trajectories in a variety of scenarios. Systems and methods described herein use deep neural networks to quickly predict optimized robotic arm trajectories according to certain constraints. Optimization, in accordance with some embodiments of the present technology, may include optimizing trajectory geometry and dynamics while satisfying a number of constraints, including staying collision-free, and minimizing the time it takes to complete the task.
-
公开(公告)号：US20210229292A1
公开(公告)日：2021-07-29
申请号：US17161297
申请日：2021-01-28
Applicant: Embodied Intelligence Inc.
Inventor： YuXuan Liu , Xi Chen , Nikhil Mishra
Abstract: Various embodiments of the present technology generally relate to robotic devices and artificial intelligence. More specifically, some embodiments relate to modeling uncertainty in neural network predictions using bounding box predictions for imaged objects. In some embodiments, a computer vision system for guiding robotic picking utilizes a method for uncertainty modeling that comprises identifying a three-dimensional object in one or more images of a scene, wherein at least one side of the 3D object is not visible to the computer vision system. The method further comprises predicting a plurality of volumes that comprise the object, wherein each volume of the plurality of volumes comprises at least a portion of the object. From the plurality of volumes, a confidence level may be determined for each volume, wherein the confidence level represents a likelihood that the volume contains the entire object.
-
-
-
-
-
-
-
-
-
Search Results
15
records
(took 0.013 seconds)
