1. 【算法+图像处理】2D卷积与快速卷积算法C语言实现
https://blog.csdn.net/guduruyu/article/details/78385418?utm_source=blogxgwz0
用“ 快速卷积”来加速‘卷积’ (不同放大倍率的VDSR 重建程序)https://blog.csdn.net/juebai123/article/details/80960654?utm_source=blogxgwz2
2. 一种快速卷积实现方法(FFT变换)
https://blog.csdn.net/B1009/article/details/78922764?utm_source=blogxgwz4
利用FFT实现快速卷积实验https://download.csdn.net/download/qq_27975371/8800473
挺有意思的一篇扫盲文,当年学校,深有同感:卷积与傅里叶变换https://blog.csdn.net/Augusdi/article/details/12438011?utm_source=blogxgwz2
3. 高效卷积算法(综述总结性)
https://blog.csdn.net/fupotui7870/article/details/79946990?utm_source=blogxgwz0
4. More is Less——卷积网络加速
https://blog.csdn.net/shuzfan/article/details/70172346?utm_source=blogxgwz0
5. C++的矩阵乘向量加速trick(循环展开)
https://blog.csdn.net/jacke121/article/details/65440754?utm_source=blogxgwz8
6. MEC —— 优化内存与速度的卷积计算
https://blog.csdn.net/shuzfan/article/details/77427979?utm_source=blogxgwz6
7.【算法导论】矩阵乘法strassen算法
CPU上的算法复杂度从3次方到2.8次方
https://blog.csdn.net/zhuangxiaobin/article/details/36476769?utm_source=blogxgwz0
8.Sparse-Winograd CNN——权重剪枝与Winograd的结合
https://blog.csdn.net/nature553863/article/details/80307277?utm_source=blogxgwz2
Paper地址:https://arxiv.org/abs/1802.06367
GitHub地址:https://github.com/xingyul/Sparse-Winograd-CNN
AI芯片:商汤科技基于winograd算法的FPGA方案分析
https://blog.csdn.net/evolone/article/details/80136193?utm_source=blogxgwz0
用FPGA评估卷积神经网络快速算法
https://blog.csdn.net/XingpengLu/article/details/80919665?utm_source=blogxgwz6
fpga 实现cnn
https://github.com/doonny/PipeCNN/tree/master/documents
如何简单快捷地用小型Xiliinx FPGA加速卷积神经网络CNN
https://blog.csdn.net/awai54st/article/details/78177351?utm_source=blogxgwz1
https://github.com/awai54st/PYNQ-Classification
C++性能优化技术导论
https://blog.csdn.net/zhangxinrun/article/details/7634999