Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020062054 - DATA PROCESSING METHOD AND DEVICE, AND UNMANNED AERIAL VEHICLE

Publication Number WO/2020/062054
Publication Date 02.04.2020
International Application No. PCT/CN2018/108401
International Filing Date 28.09.2018
IPC
G06N 3/04 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architecture, e.g. interconnection topology
CPC
B64D 47/08
BPERFORMING OPERATIONS; TRANSPORTING
64AIRCRAFT; AVIATION; COSMONAUTICS
DEQUIPMENT FOR FITTING IN OR TO AIRCRAFT; FLYING SUITS; PARACHUTES; ARRANGEMENTS OR MOUNTING OF POWER PLANTS OR PROPULSION TRANSMISSIONS IN AIRCRAFT
47Equipment not otherwise provided for
08Arrangements of cameras
G06N 3/02
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
G06N 3/04
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architectures, e.g. interconnection topology
H04N 7/18
HELECTRICITY
04ELECTRIC COMMUNICATION TECHNIQUE
NPICTORIAL COMMUNICATION, e.g. TELEVISION
7Television systems
18Closed circuit television systems, i.e. systems in which the signal is not broadcast
Applicants
  • 深圳市大疆创新科技有限公司 SZ DJI TECHNOLOGY CO., LTD. [CN]/[CN]
Inventors
  • 麻军平 MA, Junping
  • 张强 ZHANG, Qiang
  • 曹子晟 CAO, Zisheng
  • 杨康 YANG, Kang
Agents
  • 北京同立钧成知识产权代理有限公司 LEADER PATENT & TRADEMARK FIRM
Priority Data
Publication Language Chinese (ZH)
Filing Language Chinese (ZH)
Designated States
Title
(EN) DATA PROCESSING METHOD AND DEVICE, AND UNMANNED AERIAL VEHICLE
(FR) PROCÉDÉ ET DISPOSITIF DE TRAITEMENT DE DONNÉES, ET VÉHICULE AÉRIEN SANS PILOTE
(ZH) 数据处理方法、设备以及无人机
Abstract
(EN)
Embodiments of the present invention provide a data processing method and device, and an unmanned aerial vehicle. The method comprises: reading a compressed neural network parameter from a memory; decompressing the compressed neural network parameter to generate a decompressed neural network parameter; and processing data to be processed according to the decompressed neural network parameter. Since a neural network parameter is stored in a memory after being compressed, the storage space occupied by the neural network parameter is reduced, the access pressure of the memory is reduced, and the occupied bandwidth during reading of the memory is reduced.
(FR)
Des modes de réalisation de l'invention concernent un procédé et un dispositif de traitement de données, ainsi qu'un véhicule aérien sans pilote. Le procédé consiste à : lire un paramètre de réseau neuronal compressé à partir d'une mémoire ; décompresser le paramètre de réseau neuronal compressé afin de générer un paramètre de réseau neuronal décompressé ; et traiter les données à traiter selon le paramètre de réseau neuronal décompressé. Comme un paramètre de réseau neuronal est stocké dans une mémoire après avoir été compressé, l'espace de stockage occupé par le paramètre de réseau neuronal est réduit, la pression d'accès de la mémoire est réduite, et la largeur de bande occupée pendant la lecture de la mémoire est réduite.
(ZH)
本发明实施例提供一种数据处理方法、设备以及无人机,此方法包括:从存储器中读取经压缩的神经网络参数;对所述经压缩的神经网络参数进行解压,以生成解压后的神经网络参数;以及根据所述解压后的神经网络参数对待处理数据进行处理。由于神经网络参数是由经压缩后的形式存储在存储器中,减少了神经网络参数占用的存储空间,减少了存储器的访问压力,减少了读取存储器时占用的带宽。
Also published as
CN201880031720.0
Latest bibliographic data on file with the International Bureau