Object recognition apparatus for an autonomous vehicle and an object recognition method therefor
AssigneesHyundai Motor Company, Kia Corporation
InventorJong Hyun Choi
Life of this patent
Estimating the term…
Abstract
An object recognition apparatus and method for an autonomous vehicle are disclosed. The object recognition apparatus includes a processor and storage. The processor: trains a first network model based on an image; corrects distortion and rotation of the image; generates at least one distorted image based on the corrected image; trains a feature transfer module based on a feature extracted from the corrected image and a feature extracted from the at least one distorted image; inserts the feature transfer module into the first network model; and performs fine-tuning for a second network model including the feature transfer module based on the at least one distorted image.
Claims20
An object recognition apparatus for an autonomous vehicle, the object recognition apparatus comprising:
- a processor;
- and
- storage storing a deep-learning network model for performing object recognition based on an image,
- wherein the processor is configured to
- train a first network model based on an image,
- correct a distortion and a rotation of the image,
- generate at least one distorted image based on the corrected image,
- train a feature transfer module based on a feature extracted from the corrected image and a feature extracted from the at least one distorted image,
- insert the feature transfer module into the first network model, and
- perform fine-tuning for a second network model including the feature transfer module based on the at least one distorted image.
An object recognition method performed by an object recognition apparatus for an autonomous vehicle, the object recognition method comprising:
- training a first network model based on an image,
- correcting distortion and rotation of the image;
- generating at least one distorted image based on the corrected image;
- training a feature transfer module based on a feature extracted from the corrected image and a feature extracted from the at least one distorted image;
- inserting the feature transfer module into the first network model;
- and
- performing fine-tuning for a second network model including the feature transfer module based on the at least one distorted image,
- wherein each of the first network model and the second network model are a deep-learning network model for performing object recognition based on an image.
Dependencies are inferred from the claim text.