Learning method and learning device for integrating object detection information acquired through V2V communication from other autonomous vehicle with object detection information generated by present autonomous vehicle, and testing method and testing device using the same

• US 10,817,777 B2
• Filed: 12/22/2019
• Issued: 10/27/2020
• Est. Priority Date: 01/31/2019
• Status: Active Grant

• Current Assignee: Stradvision, Inc.
• Original Assignee: Stradvision, Inc.
• Inventors: Kim, Kye-Hyeon, Kim, Yongjoong, Kim, Hak-Kyoung, Nam, Woonhyun, Boo, SukHoon, Sung, Myungchul, Shin, Dongsoo, Yeo, Donghun, Ryu, Wooju, Lee, Myeong-Chun, Lee, Hyungsoo, Jang, Taewoong, Jeong, Kyungjoong, Je, Hongmo, Cho, Hojin
• Primary Examiner(s): Huang, Miranda M
• Assistant Examiner(s): Hinckley, Chase P.
• Application Number:
  US16/724,302
• Publication Number:
  US 20200250514A1
• Time in Patent Office: 310 Days
• Field of Search: 706 20
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06F 18/24 Classification techniques
• CPC Class Codes: G06F 18/251 of input or preprocessed data
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06V 10/25 Determination of region of interest [ROI] or a volume of interest [VOI]
• CPC Class Codes: G06V 10/774 Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06V 10/776 Validation; Performance evaluation
• CPC Class Codes: G06V 10/803 of input or preprocessed data
• CPC Class Codes: G06V 10/82 using neural networks
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads

Abstract

A learning method for generating integrated object detection information by integrating first object detection information and second object detection information is provided. And the method includes steps of: (a) a learning device instructing a concatenating network to generate one or more pair feature vectors; (b) the learning device instructing a determining network to apply FC operations to the pair feature vectors, to thereby generate (i) determination vectors and (ii) box regression vectors; (c) the learning device instructing a loss unit to generate an integrated loss by referring to the determination vectors, the box regression vectors and their corresponding GTs, and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN.

Claims

1. A learning method for generating integrated object detection information on an integrated target space including a first target space and a second target space, by integrating first object detection information on the first target space generated by a first vehicle and second object detection information on the second target space generated by a second vehicle, comprising steps of:

(a) a learning device, if the first object detection information on the first target space and the second object detection information on the second target space are acquired by processing a first original image on the first target space and a second original image on the second target space, instructing a concatenating network included in a Deep Neural Network (DNN) to generate one or more pair feature vectors including information on one or more pairs of first original Regions-of-Interest (ROIs) included in the first target space and second original ROIs in the second target space;

(b) the learning device instructing a determining network included in the DNN to apply one or more fully-connected operations to the pair feature vectors, to thereby generate (i) one or more determination vectors including information on probabilities of the first original ROIs and the second original ROIs included in each of the pairs being appropriate to be integrated and (ii) one or more box regression vectors including information on each of relative 3-dimensional locations of integrated ROIs, corresponding to at least part of the pairs, comparing to each of original 3-dimensional locations of each component of said at least part of the pairs, on the integrated target space;

(c) the learning device instructing a loss unit to generate an integrated loss by referring to the determination vectors, the box regression vectors and their corresponding Ground Truths (GTs), and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN, wherein

at the step of (a), a specific pair feature vector, which is one of the pair feature vectors, includes (i) first class information of a first specific object included in the first target space, (ii) feature values of a first specific original ROI including the first specific object, (iii) 3-dimensional coordinate values of a first specific original bounding box corresponding to the first specific original ROI, (iv) 3-dimensional coordinate values of the first specific original ROI, (v) second class information of a second specific object included in the second target space, (vi) feature values of a second specific original ROI including the second specific object, and (vii) 3-dimensional coordinate values of a second specific original bounding box corresponding to the second specific original ROI, and (viii) 3-dimensional coordinate values of the second specific original ROI, and

at the step of (b), a specific determination vector, which is one of the determination vectors and corresponds to the specific pair feature vector, includes information on a probability of the first specific original ROI and the second specific original ROI being integrated on the integrated target space, and a specific box regression vector, which is one of the box regression vectors and corresponds to the specific pair feature vector, includes information on 3-dimensional coordinates of a specific integrated bounding box generated by merging the first specific original ROI and the second specific original ROI on the integrated target space.

2. The method of claim 1, wherein, at the step of (c), the learning device instructs the loss unit (i) to generate a determination loss by using at least part of the determination vectors through a cross entropy method, (ii) to generate a box regression loss by using at least part of the box regression vectors through a smooth-L1 method, and (iii) to generate the integrated loss by referring to the determination loss and the box regression loss.

3. The method of claim 2, wherein, at the step of (c), the determination loss is generated by a formula below:

4. The method of claim 1, wherein the learning device instructs each of deep learning neurons included in one or more layers of the DNN to repeatedly apply one or more convolutional operations to its input by using its own at least one parameter and deliver its output to its next deep learning neuron, to thereby generate the pair feature vectors, the determination vectors and the box regression vectors.

5. The method of claim 1, wherein, at the step of (b), the learning device instructs the determining network included in the DNN to generate the determination vectors by applying at least part of the FC operations to the pair feature vectors, and to generate the one or more box regression vectors corresponding to one or more specific pair feature vectors, among the pair feature vectors, whose values in corresponding specific determination vectors denoting specific probabilities of specific pairs to be integrated are larger than a prescribed threshold, by applying any part of the FC operations to the specific pair feature vectors.

6. A testing method for generating integrated object detection information for testing on an integrated target space for testing including a first target space for testing and a second target space for testing, by integrating first object detection information for testing on the first target space for testing generated by a first vehicle for testing and second object detection information for testing on the second target space for testing generated by a second vehicle for testing, comprising steps of:

(a) on condition that (1) a learning device, if first object detection information for training on a first target space for training and second object detection information for training on a second target space for training have been acquired by processing a first original image for training on the first target space for training and a second original image for training on the second target space for training, has instructed a concatenating network included in a Deep Neural Network (DNN) to generate one or more pair feature vectors for training including information on one or more pairs for training of first original Regions-of-Interest (ROIs) for training included in the first target space for training and second original ROIs for training in the second target space for training;

(2) the learning device has instructed a determining network included in the DNN to apply one or more fully-connected operations to the pair feature vectors for training, to thereby generate (i) one or more determination vectors for training including information on probabilities for training of the first original ROIs for training and the second original ROIs for training included in each of the pairs for training being appropriate to be integrated and (ii) one or more box regression vectors for training including information on each of relative 3-Dimensional locations for training of integrated ROIs for training, corresponding to at least part of the pairs for training, comparing to each of original 3-Dimensional locations for training of each component of said at least part of the pairs for training, on an integrated target space for training;

(3) the learning device has instructed a loss unit to generate an integrated loss by referring to the determination vectors for training, the box regression vectors for training and their corresponding Ground Truths (GTs), and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN, a testing device installed on the first vehicle, if the first object detection information for testing on the first target space for testing and the second object detection information for testing on the second target space for testing are acquired by processing a first original image for testing on the first target space for testing and a second original image for testing on the second target space for testing, instructing the concatenating network included in the DNN to generate one or more pair feature vectors for testing including information on one or more pairs for testing of first original ROIs for testing included in the first target space for testing and second original ROIs for testing in the second target space for testing;

(b) the testing device instructing the determining network included in the DNN to apply the FC operations to the pair feature vectors for testing, to thereby generate (i) one or more determination vectors for testing including information on probabilities for testing of the first original ROIs for testing and the second original ROIs for testing included in each of the pairs for testing being appropriate to be integrated and (ii) one or more box regression vectors for testing including information on each of relative 3-dimensional locations for testing of integrated ROIs for testing, corresponding to at least part of the pairs for testing, comparing to each of original 3-Dimensional locations for testing of each component of said at least part of the pairs for testing, on the integrated target space for testing;

(c) the testing device instructing a merging unit to generate the integrated object detection information for testing by merging at least part of the pairs for testing of first original bounding boxes for testing and second original bounding boxes for testing by referring to the determination vectors for testing and the box regression vectors for testing.

7. The method of claim 6, wherein the first original image for testing on the first target space for testing, acquired through at least part of one or more first cameras, one or more first lidars and one or more first radars installed on the first vehicle, is processed by a first neural network included in the first vehicle, to thereby generate the first object detection information for testing including (i) first class information for testing on objects for testing included in the first target space for testing, (ii) feature values for testing of the first original ROIs for testing, (iii) 3-dimensional coordinate values for testing of the first original bounding boxes for testing and (iv) 3-dimensional coordinate values for testing of the first original ROIs for testing, and

the second original image for testing on the second target space for testing, acquired through at least part of one or more second cameras, one or more second lidars and one or more second radars installed on the second vehicle, is processed by a second neural network included in the second vehicle, to thereby generate the second object detection information including (i) second class information for testing on objects for testing included in the second target space for testing, (ii) feature values for testing of the second original ROIs for testing, (iii) 3-dimensional coordinate values for testing of the second original bounding boxes for testing and (iv) 3-dimensional coordinate values for testing of the second original ROIs for testing, and the second object detection information is delivered to the first vehicle through a V2V communication.

8. The method of claim 6, wherein, at the step of (a), a specific pair feature vector for testing, which is one of the pair feature vectors for testing, includes (i) first class information for testing of a first specific object for testing included in the first target space for testing, (ii) feature values for testing of a first specific original ROI for testing including the first specific object for testing, (iii) 3-dimensional coordinate values of a first specific original bounding box for testing corresponding to the first specific original ROI for testing, (iv) 3-dimensional coordinate values of the first specific original ROI for testing, (v) second class information for testing of a second specific object for testing included in the second target space for testing, (vi) feature values for testing of a second specific original ROI for testing including the second specific object for testing, (vii) 3-dimensional coordinate values of a second specific original bounding box for testing corresponding to the second specific original ROI for testing, and (viii) 3-dimensional coordinate values of the second specific original ROI for testing.

9. The method of claim 8, wherein, at the step of (b), a specific determination vector for testing, which is one of the determination vectors for testing and corresponds to the specific pair feature vector for testing, includes information on a probability of the first specific original ROI for testing and the second specific original ROI for testing being integrated on the integrated target space for testing, and a specific box regression vector for testing, which is one of the box regression vectors for testing and corresponds to the specific pair feature vector for testing, includes information on 3-dimensional coordinates of a specific integrated bounding box for testing generated by merging the first specific original ROI for testing and the second specific original ROI for testing on the integrated target space for testing.

10. A learning device for generating integrated object detection information on an integrated target space including a first target space and a second target space, by integrating first object detection information on the first target space generated by a first vehicle and second object detection information on the second target space generated by a second vehicle, comprising:

at least one memory that stores instructions; and

at least one processor configured to execute the instructions to perform processes of (I) if the first object detection information on the first target space and the second object detection information on the second target space are acquired by processing a first original image on the first target space and a second original image on the second target space, instructing a concatenating network included in a Deep Neural Network (DNN) to generate one or more pair feature vectors including information on one or more pairs of first original Regions-of-Interest (ROIs) included in the first target space and second original ROIs in the second target space;

(II) instructing a determining network included in the DNN to apply one or more fully-connected operations to the pair feature vectors, to thereby generate (i) one or more determination vectors including information on probabilities of the first original ROIs and the second original ROIs included in each of the pairs being appropriate to be integrated and (ii) one or more box regression vectors including information on each of relative 3-dimensional locations of integrated ROIs, corresponding to at least part of the pairs, comparing to each of original 3-dimensional locations of each component of said at least part of the pairs, on the integrated target space;

(III) instructing a loss unit to generate an integrated loss by referring to the determination vectors, the box regression vectors and their corresponding Ground Truths (GTs), and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN, wherein

at the process of (I), a specific pair feature vector, which is one of the pair feature vectors, includes (i) first class information of a first specific object included in the first target space, (ii) feature values of a first specific original ROI including the first specific object, (iii) 3-dimensional coordinate values of a first specific original bounding box corresponding to the first specific original ROI, (iv) 3-dimensional coordinate values of the first specific original ROI, (v) second class information of a second specific object included in the second target space, (vi) feature values of a second specific original ROI including the second specific object, and (vii) 3-dimensional coordinate values of a second specific original bounding box corresponding to the second specific original ROI, and (viii) 3-dimensional coordinate values of the second specific original ROI, and

at the process of (II), a specific determination vector, which is one of the determination vectors and corresponds to the specific pair feature vector, includes information on a probability of the first specific original ROI and the second specific original ROI being integrated on the integrated target space, and a specific box regression vector, which is one of the box regression vectors and corresponds to the specific pair feature vector, includes information on 3-dimensional coordinates of a specific integrated bounding box generated by merging the first specific original ROI and the second specific original ROI on the integrated target space.

11. The device of claim 10, wherein, at the process of (III), the processor instructs the loss unit (i) to generate a determination loss by using at least part of the determination vectors through a cross entropy device, (ii) to generate a box regression loss by using at least part of the box regression vectors through a smooth-L1 device, and (iii) to generate the integrated loss by referring to the determination loss and the box regression loss.

12. The device of claim 11, wherein, at the process of (III), the determination loss is generated by a formula below:

13. The device of claim 10, wherein the processor instructs each of deep learning neurons included in one or more layers of the DNN to repeatedly apply one or more convolutional operations to its input by using its own at least one parameter and deliver its output to its next deep learning neuron, to thereby generate the pair feature vectors, the determination vectors and the box regression vectors.

14. The device of claim 10, wherein, at the process of (II), the processor instructs the determining network included in the DNN to generate the determination vectors by applying at least part of the FC operations to the pair feature vectors, and to generate the one or more box regression vectors corresponding to one or more specific pair feature vectors, among the pair feature vectors, whose values in corresponding specific determination vectors denoting specific probabilities of specific pairs to be integrated are larger than a prescribed threshold, by applying the other any part of the FC operations to the specific pair feature vectors.

15. A testing device for generating integrated object detection information for testing on an integrated target space for testing including a first target space for testing and a second target space for testing, by integrating first object detection information for testing on the first target space for testing generated by a first vehicle for testing and second object detection information for testing on the second tamet space for testing generated by a second vehicle for testing, comprising:

at least one memory that stores instructions; and

at least one processor configured to execute the instructions to perform processes of;

(I) on condition that (1) a learning device, if first object detection information for training on a first target space for training and second object detection information for training on a second target space for training have been acquired by processing a first original image for training on the first target space for training and a second original image for training on the second target space for training, has instructed a concatenating network included in a Deep Neural Network (DNN) to generate one or more pair feature vectors for training including information on one or more pairs for training of first original Regions-of-Interest (ROIs) for training included in the first target space for training and second original ROIs for training in the second target space for training;

(2) the learning device has instructed a determining network included in the DNN to apply one or more fully-connected operations to the pair feature vectors for training, to thereby generate (i) one or more determination vectors for training including information on probabilities for training of the first original ROIs for training and the second original ROIs for training included in each of the pairs for training being appropriate to be integrated and (ii) one or more box regression vectors for training including information on each of relative 3-Dimensional locations for training of integrated ROIs for training, corresponding to at least part of the pairs for training, comparing to each of original 3-Dimensional locations for training of each component of said at least part of the pairs for training, on an integrated target space for training;

(3) the learning device has instructed a loss unit to generate an integrated loss by referring to the determination vectors for training, the box regression vectors for training and their corresponding Ground Truths (GTs), and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN, if the first object detection information for testing on the first target space for testing and the second object detection information for testing on the second target space for testing are acquired by processing a first original image for testing on the first target space for testing and a second original image for testing on the second target space for testing, instructing the concatenating network included in the DNN to generate one or more pair feature vectors for testing including information on one or more pairs for testing of first original ROIs for testing included in the first target space for testing and second original ROIs for testing in the second target space for testing;

(II) instructing the determining network included in the DNN to apply the FC operations to the pair feature vectors for testing, to thereby generate (i) one or more determination vectors for testing including information on probabilities for testing of the first original ROIs for testing and the second original ROIs for testing included in each of the pairs for testing being appropriate to be integrated and (ii) one or more box regression vectors for testing including information on each of relative 3-dimensional locations for testing of integrated ROIs for testing, corresponding to at least part of the pairs for testing, comparing to each of original 3-Dimensional locations for testing of each component of said at least part of the pairs for testing, on the integrated target space for testing;

(III) instructing a merging unit to generate the integrated object detection information for testing by merging at least part of the pairs for testing of first original bounding boxes for testing and second original bounding boxes for testing by referring to the determination vectors for testing and the box regression vectors for testing.

16. The device of claim 15, wherein the first original image for testing on the first target space for testing, acquired through at least part of one or more first cameras, one or more first lidars and one or more first radars installed on the first vehicle, is processed by a first neural network included in the first vehicle, to thereby generate the first object detection information for testing including (i) first class information on objects for testing included in the first target space for testing, (ii) feature values for testing of the first original ROIs for testing, (iii) 3-dimensional coordinate values for testing of the first original bounding boxes for testing and (iv) 3-dimensional coordinate values for testing of the first original ROIs for testing, and

the second original image for testing on the second target space for testing, acquired through at least part of one or more second cameras, one or more second lidars and one or more second radars installed on the second vehicle, is processed by a second neural network included in the second vehicle, to thereby generate the second object detection information including (i) second class information on objects for testing included in the second target space for testing, (ii) feature values for testing of the second original ROIs for testing, (iii) 3-dimensional coordinate values for testing of the second original bounding boxes for testing and (iv) 3-dimensional coordinate values for testing of the second original ROIs for testing, and the second object detection information is delivered to the first vehicle through a V2V communication.

17. The device of claim 15, wherein, at the process of (I), a specific pair feature vector for testing, which is one of the pair feature vectors for testing, includes (i) first class information for testing of a first specific object for testing included in the first target space for testing, (ii) feature values for testing of a first specific original ROI for testing including the first specific object for testing, (iii) 3-dimensional coordinate values of a first specific original bounding box for testing corresponding to the first specific original ROI for testing, (iv) 3-dimensional coordinate values of the first specific original ROI for testing, (v) second class information for testing of a second specific object for testing included in the second target space for testing, (vi) feature values for testing of a second specific original ROI for testing including the second specific object for testing, (vii) 3-dimensional coordinate values of a second specific original bounding box for testing corresponding to the second specific original ROI for testing, and (viii) 3-dimensional coordinate values of the second specific original ROI for testing.

18. The device of claim 17, wherein, at the process of (II), a specific determination vector for testing, which is one of the determination vectors for testing and corresponds to the specific pair feature vector for testing, includes information on a probability of the first specific original ROI for testing and the second specific original ROI for testing being integrated on the integrated target space for testing, and a specific box regression vector for testing, which is one of the box regression vectors for testing and corresponds to the specific pair feature vector for testing, includes information on 3-dimensional coordinates of a specific integrated bounding box for testing generated by merging the first specific original ROI for testing and the second specific original ROI for testing on the integrated target space for testing.