cvnn
Library to help implement a complex-valued neural network (cvnn) using tensorflow as back-end
Senior ML Engineer · LLM systems & agents in production · Ph.D. CentraleSupélec · 150k+ pip downloads · Paris
Library to help implement a complex-valued neural network (cvnn) using tensorflow as back-end
This project aim is to use the Cypress CYUSB3KIT-003 EZ-USB FX3 SuperSpeed Explorer Kit to both program and communicate with a Xilinx Spartan 6 FPGA embedded on the SP605 Evaluation Kit. In order to connect both boards, the CYUSB3ACC-005 FMC Interconnect Board was used.
Core code for simulations used for most of my publications
PolSAR classification / segmentation using complex-valued neural networks.
The known game Forbidden Desert for PC for two players (LAN mode available)
The aim of this project was to speed-up a target application through multiple parallelism models (MPI, OpenMP and CUDA) for various architectures (CPU and GPU) and conclude which approach is the most suitable for the specific application.
A compilator was created to generate a x86-64 assembler code from a C fragment called mini-C. This is a 100% C-compatible fragment, in the sense that any Mini C program is also a C program.