Processing

Please wait...

Settings

Settings

Goto Application

1. WO2018136676 - MULTI-VIEW EMBEDDING WITH SOFT-MAX BASED COMPATIBILITY FUNCTION FOR ZERO-SHOT LEARNING

Publication Number WO/2018/136676
Publication Date 26.07.2018
International Application No. PCT/US2018/014302
International Filing Date 19.01.2018
IPC
G06K 9/20 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
20Image acquisition
G06K 9/00 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
G06K 9/48 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
36Image preprocessing, i.e. processing the image information without deciding about the identity of the image
46Extraction of features or characteristics of the image
48by coding the contour of the pattern
G06N 3/08 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
CPC
G05D 1/0088
GPHYSICS
05CONTROLLING; REGULATING
DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
1Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
0088characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours
G05D 1/0246
GPHYSICS
05CONTROLLING; REGULATING
DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
1Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
02Control of position or course in two dimensions
021specially adapted to land vehicles
0231using optical position detecting means
0246using a video camera in combination with image processing means
G06K 9/00684
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
00624Recognising scenes, i.e. recognition of a whole field of perception; recognising scene-specific objects
00664Recognising scenes such as could be captured by a camera operated by a pedestrian or robot, including objects at substantially different ranges from the camera
00684Categorising the entire scene, e.g. birthday party or wedding scene
G06K 9/46
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
36Image preprocessing, i.e. processing the image information without deciding about the identity of the image
46Extraction of features or characteristics of the image
G06K 9/6232
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
6217Design or setup of recognition systems and techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
6232Extracting features by transforming the feature space, e.g. multidimensional scaling; Mappings, e.g. subspace methods
G06K 9/6256
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
6217Design or setup of recognition systems and techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
6256Obtaining sets of training patterns; Bootstrap methods, e.g. bagging, boosting
Applicants
  • HRL LABORATORIES, LLC [US]/[US]
Inventors
  • CAO, Xi, Hang
  • KIM, Kyungnam
Agents
  • TOPE-MCKAY, Cary, R.
Priority Data
15/874,85618.01.2018US
62/448,38219.01.2017US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) MULTI-VIEW EMBEDDING WITH SOFT-MAX BASED COMPATIBILITY FUNCTION FOR ZERO-SHOT LEARNING
(FR) INTÉGRATION DE MULTIPLES VUES À FONCTION DE COMPATIBILITÉ SOFT-MAX D'APPRENTISSAGE SANS EXEMPLE
Abstract
(EN)
Described is a system for multi-view embedding for object recognition. Daring operation, the system receives an input image and transforms raw data of objects in the image with corresponding labels into low-level features and high-level semantic representations of the labels. A trained object recognition model is generated by embedding the low-level features with multiple high-level semantic representations. The system then receives data of an unknown object and assigns a label to the unknown object using the trained object recognition model. Finally, a device can be controlled based on the label.
(FR)
L'invention concerne un système d'intégration de multiples vues de reconnaissance d'objet. En fonctionnement, le système reçoit une image d'entrée et transforme des données brutes d'objets de l'image à étiquettes correspondantes en caractéristiques de niveau bas et en représentations sémantiques de niveau haut des étiquettes. Un modèle de reconnaissance d'objet entraîné est généré par intégration, dans les caractéristiques de niveau bas, de multiples représentations sémantiques de niveau haut. Le système reçoit ensuite des données d'un objet inconnu et attribue une étiquette à l'objet inconnu à l'aide du modèle de reconnaissance d'objet entraîné. Enfin, un dispositif peut être commandé sur la base de l'étiquette.
Also published as
Latest bibliographic data on file with the International Bureau