专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明公开

EP4303835A1 ANNOTATION OF OBJECTS IN IMAGE FRAMES 审中-公开
公开(公告)号：EP4303835A1
公开(公告)日：2024-01-10
申请号：EP22183357.7
申请日：2022-07-06
申请人： Zenseact AB
发明人： VERBEKE, Willem , LJUNGBERGH, William , MÅNSSON, Olle
IPC分类号： G06V20/58
摘要： The present disclosure relates to a method performed by an annotation system (1) for supporting annotation of objects in image frames of a traffic environment-related video sequence. The annotation system determines (1001) an annotation of an object (4) in an image frame (fo) of the video sequence, which annotation comprises at least a first property (41) of the object; tracks (1002) the object through the video sequence; and assigns (1003) the at least first object property to the object (4') in one or more previous and/or subsequent image frames (f p/s ) of the video sequence. The annotation system further identifies (1004) at least a first of the previous and/or subsequent image frames (f p/s ) in which: pixel area dimensions (5) of the object (4') are below an object type- and/or property type-specific threshold stipulating pixel area dimensions below which the at least first object property is defined undetectable, e.g. to a human; and/or motion blur (6) in pixels of at least a predeterminable portion of the object (4') exceeds a motion blur threshold stipulating a motion blur level above which the at least first object property is defined undetectable, e.g. to a human; and/or brightness (7) in pixels of at least a predeterminable portion of the object (4') exceeds a brightness threshold stipulating a brightness level above which the at least first object property is defined undetectable, e.g. to a human. Moreover, the annotation system appoints (1005) the at least first identified image frame as annotation data.
The disclosure also relates to an annotation system in accordance with the foregoing, an arrangement - for instance an offboard system (109) and/or a vehicle (2) comprising such an annotation system, and a respective corresponding computer program product and non-volatile computer readable storage medium.

2. 发明公开

EP4394667A1 TARGETED DATA GENERATION BY NEURAL NETWORK ENSEMBLES 审中-公开
公开(公告)号：EP4394667A1
公开(公告)日：2024-07-03
申请号：EP22216937.7
申请日：2022-12-28
申请人： Zenseact AB
发明人： VERBEKE, Willem , LJUNGBERGH, William
IPC分类号： G06N20/20
CPC分类号： G06N20/20 , G06N3/09 , G06N3/045 , G06N3/098 , G06N3/0985 , G06N3/096 , G06N3/094 , G06N3/091 , B60W60/00
摘要： The present disclosure relates to methods, systems, a vehicle and a computer-readable storage medium and a computer program product. The method comprises obtaining a trained ensemble of machine learning, ML, algorithms, comprising a plurality of ML algorithms that are trained at least partly based on a first set of training data. The method further comprises forming an input data set for the trained ensemble of ML algorithms. The method further comprises providing the formed input data set to the trained ensemble of ML algorithms for generating, for each of the one or more input data samples, an ensemble prediction output. In response to a determined discrepancy in the ensemble prediction output for a particular input data sample, the method further comprises forming a second training data set based on that particular input data sample.

3. 发明公开

EP4358039A1 LANE-ASSIGNMENT FOR TRAFFIC OBJECTS ON A ROAD 审中-公开
公开(公告)号：EP4358039A1
公开(公告)日：2024-04-24
申请号：EP22202398.8
申请日：2022-10-19
申请人： Zenseact AB
发明人： VERBEKE, Willem , Olle, MÅNSSON
IPC分类号： G06V20/58 , G06V20/56 , G06F18/214 , G06V20/59
CPC分类号： G06V20/58 , G06V20/582 , G06V20/588 , G06F18/214 , G06V20/597
摘要： The present disclosure relates to a method, system, a vehicle and a computer-readable storage medium. The method comprises obtaining sensor data from a sensor system of an ego vehicle comprising an Automated Driving System, ADS, and traveling on the road. The method further comprises identifying the at least one roadside traffic object in the surrounding environment of the ego vehicle and determining a change in a driving behavior of the ego vehicle and/or of at least one external vehicle being present in the surrounding environment of the ego vehicle. The method further comprises determining a co-occurrence of the identification of the at least one roadside traffic object and determination of the change in the driving behavior of the ego vehicle and/or of the at least one external vehicle. The method further comprises if the co-occurrence is determined, generating a corresponding image annotation for one or more obtained images of the at least one identified roadside traffic object.

4. 发明公开

EP4227908A1 ITERATIVE REFINEMENT OF ANNOTATED DATASETS 审中-实审
公开(公告)号：EP4227908A1
公开(公告)日：2023-08-16
申请号：EP22156360.4
申请日：2022-02-11
申请人： Zenseact AB
发明人： VERBEKE, Willem
IPC分类号： G06V10/70 , G06V10/774 , G06V10/776 , G06V10/778
摘要： The present invention relates to a method for improving annotated datasets for training machine-learning algorithms. More specifically, the present invention relates to a method employing an iterative scheme where a machine-learning algorithm is used trained on an annotated dataset, and subsequently used to evaluate the annotated dataset by feeding the annotated dataset into the machine-learning algorithm in order to extract an erroneous dataset from the input dataset, where the erroneous dataset comprises images associated with indications of annotation errors and/or missing annotations in the annotated dataset. Then, the erroneous dataset is re-annotated, and the training and evaluation steps are repeated until the number of images of the annotated dataset with annotation errors and/or missing annotations is below a threshold.

5. 发明公开

EP4394664A1 AUTOMATED DATA GENERATION BY NEURAL NETWORK ENSEMBLES 审中-公开
公开(公告)号：EP4394664A1
公开(公告)日：2024-07-03
申请号：EP22217128.2
申请日：2022-12-29
申请人： Zenseact AB
发明人： VERBEKE, Willem , LJUNGBERGH, William
IPC分类号： G06N3/094 , G06N20/20 , G06N3/045 , G06N3/09 , G06N3/098 , G06N3/091 , G06N3/096 , G06N3/0985
CPC分类号： G06N20/20 , G06N3/09 , G06N3/045 , G06N3/098 , G06N3/0985 , G06N3/096 , G06N3/094 , G06N3/091 , B60W60/00
摘要： The present disclosure relates to methods, systems, a vehicle and a computer-readable storage medium and a computer program product. The method comprises obtaining a cluster of trained ensembles of machine learning, ML, algorithms. The cluster comprises two or more ML algorithm ensembles, wherein each ML ensemble comprises a plurality of ML algorithms that are trained at least partly based on a first set of training data. The method further comprises obtaining sensor data representative of a scenario, in a surrounding environment of a vehicle observed by at least two sensor devices comprised in a sensor system of the vehicle. The sensor data comprises at least two sensor data sets. The method further comprises providing each obtained sensor data set as input to a corresponding ML algorithm ensemble being comprised in the ML algorithm ensemble cluster. The method further comprises selecting the ensemble-prediction output of at least one ML algorithm ensemble associated with an absent determined discrepancy for generating an annotation for one or more data samples of the sensor data set of at least one ML algorithm ensemble associated with a determined discrepancy.

6. 发明公开

EP4394626A1 CLASSIFICATION OF OBJECTS PRESENT ON A ROAD 审中-公开
公开(公告)号：EP4394626A1
公开(公告)日：2024-07-03
申请号：EP22216820.5
申请日：2022-12-27
申请人： Zenseact AB
发明人： VERBEKE, Willem , JOHNANDER, Joakim
IPC分类号： G06F18/21 , G06V20/58 , G06V30/19
CPC分类号： G06V30/19147 , G06F18/2185 , G06V20/58
摘要： The present disclosure relates to a method, system, a vehicle and a computer-readable storage medium and a computer program product for determining an association of at least one object to a target object class, the at least one object being present in a surrounding environment of a vehicle travelling on a road. The method comprises obtaining sensor data from a sensor system of the vehicle, the sensor data comprising one or more images, captured by a vehicle-mounted camera, of the surrounding environment of the vehicle. The method further comprises determining a presence of the at least one object in the surrounding environment of the vehicle based on the obtained sensor data and obtaining a finite sub-set of annotated images being representative of at least the target object class. The finite sub-set comprises a number of annotated images smaller than a threshold value number. Further, the method comprises determining a preliminary association between the at least one object and the target object class and when the preliminary association is determined, the method further comprises producing an image label for the one or more images of the at least one object.

7. 发明公开

EP4357944A1 IDENTIFICATION OF UNKNOWN TRAFFIC OBJECTS 审中-公开
公开(公告)号：EP4357944A1
公开(公告)日：2024-04-24
申请号：EP22202765.8
申请日：2022-10-20
申请人： Zenseact AB
发明人： VERBEKE, Willem , MÅNSSON, Olle
IPC分类号： G06F18/214 , G06V20/58 , G06V40/20
CPC分类号： G06V20/58 , G06F18/214 , G06F18/2148 , G06V40/20
摘要： The present disclosure relates to a method, system, a vehicle and a computer-readable storage medium for generating training data for a machine learning, ML, algorithm configured for identification of at least one unknown traffic object on a road. The method comprises obtaining sensor data from a sensor system of an ego vehicle travelling on the road, the sensor data comprising one or more images, captured by a vehicle-mounted camera, of a surrounding environment of the vehicle and information indicative of a speed of the ego vehicle and/or speed of at least one external vehicle. The method further comprises determining a presence of at least one unknown traffic object in the surrounding environment of the ego vehicle and determining a change of speed of the ego vehicle and/or of the at least one external vehicle. In an instance of a co-occurrence of the determined change of speed and the determined presence of the at least one unknown traffic object, the method further comprises selecting one or more images of the at least one unknown traffic object.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式