DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Yoo, Hoi-Jun | - |
dc.contributor.advisor | 유회준 | - |
dc.contributor.author | Li, Zhiyong | - |
dc.date.accessioned | 2022-04-27T19:31:26Z | - |
dc.date.available | 2022-04-27T19:31:26Z | - |
dc.date.issued | 2021 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=963442&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/296023 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2021.8,[iii, 19 p. :] | - |
dc.description.abstract | A Hybrid floating-point (FP) and fixed-point (FXP) deep learning processor with an outlier-aware channel splitting algorithm is proposed for image-to-image applications on mobile devices. Since the high quality of the reconstructed image through deep learning based image-to-image application requires high bit-precision (> FP16), the mobile processor suffers from the high computation power and large external memory access (EMA). In this work, the proposed algorithm reduces 16-bit FP data to 8-bit FXP data, and only few outliers (< 10%) are computed in 16-bit FP while maintaining the image reconstruction quality. Therefore, it reduces EMA by 45.5%. Moreover, the hierarchical processor accelerates these dense 8-bit FXP data and sparse 16-bit FP data, and the functional L2 memory aggregates the convolution output of them by forming the pipeline, which reduces 98% of latency. The proposed system is simulated in 28nm COMS technology, and it occupies 4.16mm2. The hierarchical processor successfully demonstrates the × 4 scale Full-HD super-resolution generation achieving 76 frames-per-second (fps) with 133.3 mW power-consumption at 0.9 V supply and 3.6 TOPS/W of energy-efficiency which is × 3.27 higher than the previous 16-bit FXP processor. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Convolutional neural network (CNN)▼aoutlier-aware▼achannel split▼ahybrid precision▼aimage-to-image▼asuper-resolution▼astyle transfer▼anon-sparse CNN | - |
dc.subject | 합성 곱 네트워크▼a이상치 보정▼a채널 분할▼a하이브리드 정밀도▼a이미지-이미지간 변환▼a초 해상도 알고리즘▼a스타일 변환▼a비 희소성 합성 곱 신경망 | - |
dc.title | (A) hybrid FP-FXP precision deep learning processor with outlier compensation for image-to-image application | - |
dc.title.alternative | 이미지 변환을 위한 이상치 보정을 적용한 하이브리드 부동-고정점 소수점 딥러닝 프로세서 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전기및전자공학부, | - |
dc.contributor.alternativeauthor | 이지용 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.