DOI QR코드

DOI QR Code

Implementation of handwritten digit recognition CNN structure using GPGPU and Combined Layer

GPGPU와 Combined Layer를 이용한 필기체 숫자인식 CNN구조 구현

  • 이상일 (서경대학교 전자컴퓨터공학과) ;
  • 남기훈 (서경대학교 컴퓨터공학과) ;
  • 정준모 (서경대학교 전자공학과)
  • Received : 2017.09.13
  • Accepted : 2017.10.10
  • Published : 2017.11.30

Abstract

CNN(Convolutional Nerual Network) is one of the algorithms that show superior performance in image recognition and classification among machine learning algorithms. CNN is simple, but it has a large amount of computation and it takes a lot of time. Consequently, in this paper we performed an parallel processing unit for the convolution layer, pooling layer and the fully connected layer, which consumes a lot of handling time in the process of CNN, through the SIMT(Single Instruction Multiple Thread)'s structure of GPGPU(General-Purpose computing on Graphics Processing Units).And we also expect to improve performance by reducing the number of memory accesses and directly using the output of convolution layer not storing it in pooling layer. In this paper, we use MNIST dataset to verify this experiment and confirm that the proposed CNN structure is 12.38% better than existing structure.

CNN(Convolutional Nerual Network)는 기계학습 알고리즘 중에서도 이미지의 인식과 분류에 뛰어난 성능을 보이는 알고리즘 중 하나이다. CNN의 경우 간단하지만 많은 연산량을 가지고 있어 많은 시간이 소요된다. 따라서 본 논문에서는 CNN 수행과정에서 많은 처리시간이 소모되는 convolution layer와 pooling layer, fully connected layer의 연산수행을 SIMT(Single Instruction Multiple Thread)구조의 GPGPU(General-Purpose computing on Graphics Processing Units)를 통하여 병렬로 연산처리를 수행했다. 또한 convolution layer의 출력을 저장하지 않고 pooling layer의 입력으로 바로 사용함으로 메모리 접근횟수를 줄여 성능 향상을 기대했다. 본 논문에서는 이 실험검증을 위하여 MNIST 데이터 셋을 사용하였고 이를 통하여 제안하는 CNN 구조가 기존의 구조보다 12.38% 더 좋은 성능을 보임을 확인했다.

Keywords

References

  1. Lee Wan-Joo, "A Study on Implementation of Vehicle Number Plate Recognition Moduler" BUll Nat. Sci, Yong-In Univ, vol.8, no.2, pp.153-159, Feb. 2004.
  2. KimYeonGyu, "Improvement of Korea characters Recognition Performance Using CNN and Feature Extraction", Master thesis, Busan University, 2017
  3. Lee Seonguk, Byeon Gibeom, Kim Kisu, Hong Jiman, "Research of accelerating method of video quality measurement program using GPGPU", Smart Media Journal, vol.5, no4, pp. 66-74, Dec. 2016.
  4. Jeon Sanghui ,Development of multi-pedestrain gender classification algorithm using GPGPGU, Master thesis, KyungSung University, 2017.
  5. Heewon Kye and Junho Kim, "Acceleration techniques for GPGPGU-based Maximum Intensity Projection", Journal of Korea Multimedia Society, vol.1, no.8, pp981-991, 2011.
  6. Ye, Jaehyung, "Scale-invariant CNN for building and road detection from aerial images", Master thesis, Korea University, 2017.
  7. John Nickolls, William J Dally, "The GPU Computing Era", IEEE, vol.30, no.2, pp. 59-69, March-Aprill, 2010.