Details
Presenter(s)
![Shaoyi Chen Headshot](https://confcats-catavault.s3.amazonaws.com/CATAVault/ieeecass/master/files/styles/cc_user_photo/s3/user-pictures/20471_1.jpg?h=df1b6c88&itok=ngZ1wSYD)
Display Name
Shaoyi Chen
- Affiliation
-
AffiliationShanghaiTech University
- Country
-
CountryChina
Abstract
Convolutional neural network(CNN) is a major area of interest within the field of computer vision. CNN has emerged as powerful method for computer vision. However, due to the intensive computation of neural network algorithms and the massive memory footprint requirement, the forward propagation consumes considerable energy, which presents a tremendous challenge for deploying neural networks on embedded devices. Previous researchcite{nurvitadhi2017can} has established that FPGA is considered the platform of choice for CNN deployment. FPGA can generate corresponding accelerators according to low-precision integer arithmetic and give full play to the advantages of architecture.