Back to results
US 12,608,920 B2Granted
Machine learning method and machine learning apparatus for facial image blending
AssigneesFujitsu Limited, Osaka University
InventorsRyosuke Kawamura, Noriko Ikemoto Takemura, Hajime Nagahara
Life of this patent
Estimating the term…
Abstract
A storage unit stores therein first inference results respectively corresponding to a plurality of face images, which are output from a first model of outputting a first inference result in response to an input of a face image. A processing unit selects, from the plurality of face images, first and second face images whose distance calculated based their first inference results exceeds a threshold. The processing unit performs machine learning to train a second model of outputting a second inference result in response to an input of a face image, using a third face image obtained by combining the first and second face images.
Claims15
A non-transitory computer-readable storage medium
- storing a computer program that causes a computer to perform a process comprising:
- obtaining first inference results respectively corresponding to a plurality of face images, the first inference results being output from a first model of
- outputting each of the first inference results in response to an input of a corresponding one of the plurality of face images;
- selecting, from the plurality of face images, a first face image and a second face image whose distance calculated based on the first inference results corresponding to the first face image and the second face image exceeds a threshold;
- and
- performing machine learning to train a second model of outputting a second inference result in response to an
- input of each of the plurality of face images, using a third face image obtained by combining the first face image and the second face image, wherein
- each of the plurality of face images is a video capturing facial expression change,
- each of the first inference results includes a plurality of attention values, and
- each of the plurality of attention values indicates a degree of a temporal change in a facial expression in each of a plurality of frames included in a corresponding one of the plurality of face images.
7 more elements
A machine learning method comprising:
- obtaining, by a processor, first inference results respectively corresponding to a plurality of face images, the first inference results being output from a first model of outputting each of the first inference results in response to an input of a corresponding one of the plurality of face images;
- selecting, by the processor, from the plurality of face images, a first face image and a second face image whose distance calculated based on the first inference results corresponding to the first face image and the second face image exceeds a threshold;
- and
- performing, by the processor, machine learning to train a second model of outputting a second inference result in response to an input of each of the plurality of face images, using a third face image obtained by combining the
- first face image and the second face image, wherein
- each of the plurality of face images is a video capturing facial expression change,
- each of the first inference results includes a plurality of attention values, and
- each of the plurality of attention values indicates a degree of a temporal change in a facial expression in each of a plurality of frames included in a corresponding one of the plurality of face images.
5 more elements
A machine learning apparatus comprising:
- a memory that stores therein first inference results respectively corresponding to a plurality of face images, the first inference results being output from a first model of outputting each of the first inference results in response to an input of a corresponding one of the plurality of face images;
- and
- a processor coupled to the memory and the processor configured to:
- select, from the plurality of face images, a first face image and a second face image whose distance calculated based on the first inference results corresponding to the first face image and the second face image exceeds a threshold;
- and
- perform machine learning to train a second model of outputting a second inference result in response to an input of each of the plurality of face images, using a third face image obtained by combining the first face image and the second face image, wherein
- each of the plurality of face images is a video capturing facial expression change,
- each of the first inference results includes a plurality of attention values, and
- each of the plurality of attention values indicates a degree of a temporal change in a facial expression in each of a plurality of frames included in a corresponding one of the plurality of face images.
6 more elements
Dependencies are inferred from the claim text.