Seoul National University

Snu Rdb Foundation

Hybee, Inc.

Hyung Sin Kim

Seung Hwan Lee

Young Seok Kim

Yong Hwan Lee

Yong-Hwan Lee

Sae Woong Bahk

Jae Seok Bang

Sun Wook Hwang

A semi-supervised learning method for object detection in an autonomous vehicle and a device for performing semi-supervised learning for object detection in an autonomous vehicle can include receiving, by a server, no-label voxel data from a vehicle, performing, by the server, a data-based update on a server object detection model on the basis of label voxel data and the no-label voxel data, determining, by the server, a loss value on the basis of the label voxel data and the no-label voxel data, and performing, by the server, a loss-based update on the server object detection model using the loss value.

Semi-supervised learning method for object detection in autonomous vehicle and server for performing semi-supervised learning for object detection in autonomous vehicle

Provided is a method of subscribing, by a terminal, to a network as a router or an end device in a wireless network (WN) having a tree structure including routers and end devices. Each router receives, from a parent router thereof, addresses which are usable independently for child routers and child end devices thereof, and classifies the addresses into a general address and a special address. A terminal wishing to subscribe to a network may subscribe to the network by using a special address of an end device, only when the terminal is unable to subscribe to the network by using a general address of the end device, thereby reducing depletion of addresses to be used for terminal subscription and facilitating network subscription of the end device.

Method of network subscription in a wireless network

Disclosed is a self-construction system of a wireless sensor network, and a method for self-construction of a wireless sensor network using the same. Also disclosed is a method for self-construction of a cluster tree structure-based wireless sensor network (WSN). The present invention can connect devices distributed densely in a wide area to a network while maintaining the advantages of a conventional tree-structured network, such as the reduction of time required for network construction, traffic by the exchange of control command messages, and a load for a routing path search, and thus can provide a WSN having an improved self-construction performance.

Self-construction system of wireless sensor network and method for self-construction of wireless sensor network using the same

Growing community of inventors

Seoul, South Korea

Hyung Sin Kim