公开(公告)号：US20190318207A1

公开(公告)日：2019-10-17

申请号：US16447842

申请日：2019-06-20

Applicant: Waymo LLC

Inventor： Wan-Yen Lo ,  Abhijit Ogale ,  Yang Gao

IPC: G06K9/62 ,  G06K9/00 ,  G06N3/08 ,  G06N3/04 ,  G05D1/00 ,  G05D1/02

Abstract: In some implementations, an image classification system of an autonomous or semi-autonomous vehicle is capable of improving multi-object classification by reducing repeated incorrect classification of objects that are considered rarely occurring objects. The system can include a common instance classifier that is trained to identify and recognize general objects (e.g., commonly occurring objects and rarely occurring objects) as belonging to specified object categories, and a rare instance classifier that is trained to compute one or more rarity scores representing likelihoods that an input image is correctly classified by the common instance classifier. The output of the rare instance classifier can be used to adjust the classification output of the common instance classifier such that the likelihood of input images being incorrectly classified is reduced.

公开(公告)号：US10346724B2

公开(公告)日：2019-07-09

申请号：US15630275

申请日：2017-06-22

Applicant: Waymo LLC

Inventor： Wan-Yen Lo ,  Abhijit Ogale ,  Yang Gao

IPC: G06K9/00 ,  G06K9/62 ,  G06N3/04 ,  G06N3/08 ,  G05D1/00 ,  G05D1/02

Abstract: In some implementations, an image classification system of an autonomous or semi-autonomous vehicle is capable of improving multi-object classification by reducing repeated incorrect classification of objects that are considered rarely occurring objects. The system can include a common instance classifier that is trained to identify and recognize general objects (e.g., commonly occurring objects and rarely occurring objects) as belonging to specified object categories, and a rare instance classifier that is trained to compute one or more rarity scores representing likelihoods that an input image is correctly classified by the common instance classifier. The output of the rare instance classifier can be used to adjust the classification output of the common instance classifier such that the likelihood of input images being incorrectly classified is reduced.

公开(公告)号：US11093799B2

公开(公告)日：2021-08-17

申请号：US16447842

申请日：2019-06-20

Applicant: Waymo LLC

Inventor： Wan-Yen Lo ,  Abhijit Ogale ,  Yang Gao

IPC: G06K9/62 ,  G06N3/08 ,  G06N3/04 ,  G06K9/00 ,  G05D1/02 ,  G05D1/00

Abstract: In some implementations, an image classification system of an autonomous or semi-autonomous vehicle is capable of improving multi-object classification by reducing repeated incorrect classification of objects that are considered rarely occurring objects. The system can include a common instance classifier that is trained to identify and recognize general objects (e.g., commonly occurring objects and rarely occurring objects) as belonging to specified object categories, and a rare instance classifier that is trained to compute one or more rarity scores representing likelihoods that an input image is correctly classified by the common instance classifier. The output of the rare instance classifier can be used to adjust the classification output of the common instance classifier such that the likelihood of input images being incorrectly classified is reduced.

公开(公告)号：US11055577B2

公开(公告)日：2021-07-06

申请号：US16447842

申请日：2019-06-20

Applicant: Waymo LLC

Inventor： Wan-Yen Lo ,  Abhijit Ogale ,  Yang Gao

IPC: G06K9/62 ,  G06N3/08 ,  G06N3/04 ,  G06K9/00 ,  G05D1/02 ,  G05D1/00

Abstract: In some implementations, an image classification system of an autonomous or semi-autonomous vehicle is capable of improving multi-object classification by reducing repeated incorrect classification of objects that are considered rarely occurring objects. The system can include a common instance classifier that is trained to identify and recognize general objects (e.g., commonly occurring objects and rarely occurring objects) as belonging to specified object categories, and a rare instance classifier that is trained to compute one or more rarity scores representing likelihoods that an input image is correctly classified by the common instance classifier. The output of the rare instance classifier can be used to adjust the classification output of the common instance classifier such that the likelihood of input images being incorrectly classified is reduced.

公开(公告)号：US20180373963A1

公开(公告)日：2018-12-27

申请号：US15630275

申请日：2017-06-22

Applicant: Waymo LLC

Inventor： Wan-Yen Lo ,  Abhijit Ogale ,  Yang Gao

IPC: G06K9/62 ,  G05D1/02 ,  G05D1/00 ,  G06N3/08

CPC classification number: G06K9/6267 ,  G05D1/0088 ,  G05D1/0246 ,  G06K9/00805 ,  G06K9/00818 ,  G06K9/6273 ,  G06N3/0454 ,  G06N3/08 ,  G06N3/084

Abstract: In some implementations, an image classification system of an autonomous or semi-autonomous vehicle is capable of improving multi-object classification by reducing repeated incorrect classification of objects that are considered rarely occurring objects. The system can include a common instance classifier that is trained to identify and recognize general objects (e.g., commonly occurring objects and rarely occurring objects) as belonging to specified object categories, and a rare instance classifier that is trained to compute one or more rarity scores representing likelihoods that an input image is correctly classified by the common instance classifier. The output of the rare instance classifier can be used to adjust the classification output of the common instance classifier such that the likelihood of input images being incorrectly classified is reduced.