公开(公告)号：US20250166345A1

公开(公告)日：2025-05-22

申请号：US18841409

申请日：2022-03-02

Applicant: NEC Corporation

Inventor： Florian Beye ,  Hayato Itsumi ,  Charvi Vitthal ,  Koichi Nihei

IPC: G06V10/72 ,  G06V10/77 ,  G06V10/776 ,  G06V10/82 ,  H04N19/132 ,  H04N19/17 ,  H04N19/42

Abstract: A parameter set of a first machine learning model is determined such that a first loss function becomes larger. The first loss function indicates a degree of variation from a conditional confidence of a first image feature conditional on a third image feature to a conditional confidence of a second image feature conditional on the third image feature. The third image feature is an image feature for recognition of a subject extracted from an original image. The first image feature is discriminated in a specific region of the original image by using a first machine learning model. A parameter set of each of the second machine learning model and the third machine learning model are determined such that a second loss function becomes smaller.

公开(公告)号：US20230334672A1

公开(公告)日：2023-10-19

申请号：US18028708

申请日：2020-10-05

Applicant: NEC Corporation

Inventor： Hayato ITSUMI ,  Koichi Nihei ,  Yusuke Shinohara ,  Takanori Iwai ,  Florian Beye ,  Charvi Vitthal

IPC: G06T5/20 ,  G06T7/215 ,  G06T5/50

CPC classification number: G06T7/215 ,  G06T5/20 ,  G06T5/50 ,  G06T2207/10016 ,  G06T2207/30252

Abstract: A region prediction unit predicts the target region in the second image frame based on the target region detected in a first image frame and a time interval between the first and second image frames. The region prediction unit predicts the target region in the third image frame based on the target region detected in the second image frame and a time interval between the second and third image frames. A region determination unit determines, as an image-processing region, a region obtained by adding a margin to the target prediction region in the third image frame in accordance with an overlapping portion between the detected target region in the second image frame and the prediction region in the second image frame.

公开(公告)号：US12283110B2

公开(公告)日：2025-04-22

申请号：US18028486

申请日：2020-10-05

Applicant: NEC Corporation

Inventor： Hayato Itsumi ,  Yusuke Shinohara ,  Koichi Nihei ,  Takanori Iwai ,  Florian Beye ,  Charvi Vitthal

IPC: G06V20/56 ,  G06T7/00 ,  G06T7/62 ,  G06V10/25 ,  G06V10/764 ,  G06V20/52

Abstract: An information processing device includes: an image acquisition unit configured to acquire an image captured by an imaging unit mounted on a vehicle; a target detection unit configured to detect a target region including a target in the acquired image; a classification identification unit configured to identify classification including a type of the detected target and a size of the detected target region; and a region determination unit configured to determine a region obtained by adding a margin corresponding to the identified classification to the target region as an image processing region.

公开(公告)号：US11764806B2

公开(公告)日：2023-09-19

申请号：US17467282

申请日：2021-09-06

Applicant: NEC CORPORATION

Inventor： Charvi Vitthal ,  Florian Beye ,  Koichi Nihei ,  Hayato Itsumi ,  Takanori Iwai

IPC: H03M7/00 ,  H03M7/30 ,  G06F18/214

CPC classification number: H03M7/6005 ,  G06F18/214 ,  H03M7/3059 ,  H03M7/3066 ,  H03M7/6011

Abstract: A system includes a non-transitory computer readable medium configured to store instructions thereon; and a processor connected to the non-transitory computer readable medium. The processor is configured to execute the instructions for generating a mask based on received data from a sensor, wherein the mask includes a plurality of importance values, and each region of the received data is designated a corresponding importance value of the plurality of importance values. The processor is configured to execute the instructions for encoding the received data based on the mask; and transmitting the encoded data to a decoder for defining reconstructed data. The processor is configured to execute the instructions for computing a loss based on the reconstructed data, the received data and the mask. The processor is configured to execute the instructions for providing training to an encoder for encoding the received data based on the computed loss.

公开(公告)号：US20240153260A1

公开(公告)日：2024-05-09

申请号：US18280381

申请日：2021-03-09

Applicant: NEC Corporation

Inventor： Florian Beye ,  Yusuke Shinohara ,  Hayato Itsumi ,  Charvi Vitthal ,  Koichi Nihei ,  Takanori Iwai

IPC: G06V10/82 ,  G06V10/776

CPC classification number: G06V10/82 ,  G06V10/776

Abstract: A model learning device determines a first machine learning model so as to further increase a combined loss function obtained by combining: a first loss function indicating the level of change in the reliability of a second image feature in a feature region of a reconstructed image, from the reliability of a first image feature in a feature region of the original image; and a second loss function indicating the level of recognition error. In addition, the model learning device: determines, so as to further reduce the combined loss function, respective parameter sets for a second machine learning model used in the generation of compressed data, and a third machine learning model used in the generation of the reconstructed image from the compressed data; and determines a parameter set for the fourth machine learning model in common with that for the first machine learning model.

公开(公告)号：US20240062507A1

公开(公告)日：2024-02-22

申请号：US18271642

申请日：2021-01-15

Applicant: NEC Corporation

Inventor： Florian Beye ,  Yusuke Shinohara ,  Hayato Itsumi ,  Koichi Nihei ,  Charvi Vitthal ,  Takanori Iwai

IPC: G06V10/42 ,  G06V10/28

CPC classification number: G06V10/42 ,  G06V10/28

Abstract: An information processing device receives communication data based on feature data that indicates a feature of presented content of target data; restores the feature data on the basis of the received communication data; restores the target data on the basis of the restored feature data; performs a recognition process on the presented content of the target data on the basis of the restored feature data; and outputs information that indicates the presented content of the restored target data and a recognition result of the recognition process.