METHOD AND SYSTEM FOR GENERATING HIERARCHICAL DATASET FOR ARTIFICIAL INTELLIGENCE LEARNING, INCLUDING DATA ACQUISITION SITUATION INFORMATION

• Main Authors: SON, Haeng Seon, MIN, Kyoung Won, SHIM, Young Bo, KIM, Yun Jeong, LEE, Seon Young
• Format: Patent
• Language: English; French; Korean
• Published: 07.07.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE ORDIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; COUNTING; PERFORMING OPERATIONS; PHYSICS; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TOTHE CONTROL OF A PARTICULAR SUB-UNIT; TRANSPORTING; VEHICLES IN GENERAL

A method and a system for generating a hierarchical dataset for artificial intelligence learning, including data acquisition situation information are provided. A method for generating a GT dataset according to an embodiment of the present invention comprises: acquiring and storing vehicle data; acquiring and storing sensor data, which is generated by a sensor installed in a vehicle; and generating and storing situation information, which is information about a situation at the time of data acquisition. Accordingly, through a hierarchical dataset in which situation information at the time of sensor data acquisition in generation of a GT descriptor is hierarchically described in a descriptor, various situations and conditions at the time of acquisition can be easily analyzed and classified on the GT descriptor, so as to make artificial intelligence network learning efficient, and eventually enable an artificial intelligence network to have high recognition performance. L'invention concerne un procédé et un système de génération d'un ensemble de données hiérarchique permettant un apprentissage par intelligence artificielle, comprenant des informations de situation d'acquisition de données. Un procédé de génération d'un ensemble de données GT selon un mode de réalisation de la présente invention consiste : à acquérir et à stocker des données de véhicule ; à acquérir et à stocker des données de capteur, qui sont générées par un capteur installé dans un véhicule ; et à générer et à stocker des informations de situation, qui sont des informations concernant une situation au moment de l'acquisition de données. En conséquence, par l'intermédiaire d'un ensemble de données hiérarchique dans lequel des informations de situation au moment de l'acquisition de données de capteur lors de la génération d'un descripteur GT sont décrites de manière hiérarchique dans un descripteur, diverses situations et conditions au moment de l'acquisition peuvent être facilement analysées et classifiées sur le descripteur GT, de manière à rendre efficient un apprentissage par réseau d'intelligence artificielle et, éventuellement, à permettre à un réseau d'intelligence artificielle d'avoir une performance de reconnaissance élevée. 데이터 취득 상황 정보를 포함하는 인공지능 학습용 계층적 데이터셋 생성 방법 및 시스템이 제공된다. 본 발명의 실시예에 따른 GT 데이터셋 생성 방법은, 차량 데이터를 획득하여 저장하고, 차량에 설치된 센서에서 발생되는 센서 데이터를 획득하여 저장하며, 데이터 취득 당시의 상황에 대한 정보인 상황 정보를 생성하여 저장한다. 이에 의해, GT 디스크립터를 생성함에 있어 센서 데이터 취득 당시의 상황 정보를 디스크립터에 계층적으로 기술한 계층적 데이터셋을 통해, 다양한 취득 당시의 상황, 조건을 GT 기술자 상에서 용이하게 분석, 분류 가능하게 하여, 효율적으로 인공지능 네트워크를 학습시킴으로써, 궁극적으로 인공지능 네트워크가 높은 인식 성능을 갖을 수 있게 된다.