그림 1. 렌더링 결과 Fig 1. Result of Rendering
그림 2. 뉴런 모델 Fig 2. Neuron Model
그림 3. 뉴럴 네트워크 가속기의 구조 Fig 3. Structure of Neural Network Accelerator
그림 4. 입력과 커널(필터)의 곱셈 Fig 4. Multiplication of Input and kernel(filter)
그림 5. 일반적인 행렬 곱셈 구조 Fig 5. General Matrix multiplication structure
그림 6. 일반적인 행렬 곱셈 코드 Fig 6. General matrix multiplication code
그림 6. 병렬 행렬 곱셈 구조 Fig 6. Parallel matrix multiplication structure
그림 7. 병렬 핼렬 곱셈 코드 Fig 7. Parallel matrix multiplication code
그림 8. 이미지 처리 결과 Fig 8. Image of processing result
표 1. 픽셀당 연산 처리 시간 비교 Table 1. Comparison of processing time per pixel
References
- Rousselle, Fabrice, Claude Knaus, and Matthias Zwicker, "Adaptive rendering with non-local means filtering" ACM Transactions on Graphics (TOG) 2012.
- Li, Tzu-Mao, Yu-Ting Wu, and Yung-Yu Chuang, "SURE-based optimization for adaptive sampling and reconstruction", ACM Transactions on Graphics (TOG) 32.6 (2012): 194.
- Rousselle, Fabrice, Marco Manzi, and Matthisa Zwicker, "Robust denising usig feature and color inforamtion", Computer Graphics Forum, Vol. 32, No. 7, 2013.
- Sen, Pradeep, and Soheil Darabi, "On filtering the noise from the random parameters in Monte Carlo rendering", ACM Trans, Graph, Vol.31, No.3, 2012.
- Sangil Lee, Kihun Nam, Junmo Jung, "Implementation of handwritten digit recognition CNN structure using GPGPU and Combined Layer", JCCT, Vol.3, No4, pp. 165-169, Nov. 2017.
- Kihun Nam, "Implementation of Neural Network Accelerator for Rendering noise Reduction", ikeee, Vol.21, No4, pp. 420-425, Dec. 2017.
- Aravind Vasudevan, Andrew Anderson, David Gregg, "Parallel Multi Channel convolution using General Matrix Multiplication", In 28th IEEE International Conference on Application-specific Systems, Architectures and Processors, ASAP 2017, pp. 19-24, July. 2017.
- K. Matsumo, N. Nakastio, and S.G. Sedukhin, "Performance Tuninf of Matrix Multiplication in OpenCL on Different GPUs and CPUs", In SC Companion;High Performance Computing, Netwoking Storage and Analysis. IEEE, pp. 396-405, 2012.
- Kwang Nin Nam, Yong Jin Jeong, "Cascade CNN with CPU-FPGA Architecture for Real time Face Detection", JIKEE, Vol.21 No.4, pp. 388-396, Dec. 2017.