Browse

Structural Improvements of Single Stage Networks for Object Detection : 객체 검출을 위한 단일 단계 신경망의 구조적 개선

Cited 0 time in Web of Science Cited 0 time in Scopus
Authors
이경민
Advisor
곽노준
Major
융합과학기술대학원 융합과학부
Issue Date
2018-02
Publisher
서울대학교 대학원
Keywords
Object DetectionNeural NetworkDeep Learning
Description
학위논문 (석사)-- 서울대학교 대학원 : 융합과학기술대학원 융합과학부, 2018. 2. 곽노준.
Abstract
Recently, a lot of single stage detectors using multi-scale features have been actively proposed. They are much faster than two stage detectors that use region proposal networks (RPN) without much degradation in the detection performances. However, the feature maps in the lower layers close to the input which are responsible for detecting small objects in a single stage detector have a problem of insufficient representation power because they are too shallow. There is also a structural contradiction that the feature maps have to deliver low-level information to next layers as well as contain high-level abstraction for prediction.
In this paper, we propose a method to enrich the representation power of feature maps using Resblock and deconvolution layers. In addition, a unified prediction module is applied to generalize output results and boost earlier layersrepresentation power for prediction. The proposed method enables more precise prediction, which achieved higher score than SSD on PASCAL VOC and MS COCO. In addition, it maintains the advantage of fast computation of a single stage detector, which requires much less computation than other detectors with similar performance.
Language
English
URI
https://hdl.handle.net/10371/142273
Files in This Item:
Appears in Collections:
Graduate School of Convergence Science and Technology (융합과학기술대학원)Dept. of Transdisciplinary Studies(융합과학부)Theses (Master's Degree_융합과학부)
  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse