Summary
Chenghe 伍成和 is a founder and hands-on engineer with over a decade of R&D experience building heterogeneous compute systems that span FPGA, ARM, GPU, and full-stack web platforms. He designs and deploys computer vision and deep learning pipelines on embedded Zynq SoCs (Verilog), accelerates models on FPGA/GPU, and ships practical 3D sensing, SLAM and real-time disparity solutions used in demos and products. As founder of DeCensorMedia and 3Deepercept he blends systems-level firmware and hardware work with web-scale technologies (Django, React, Go) to create decentralized, privacy-first home/cloud IOT and stereo-vision sensor systems. His GitHub showcases low-level stereo SGM on Zynq, backpropagation experiments and a Go framework, reflecting a rare mix of algorithmic depth and production engineering. Based in Japan, he moves fluidly between FPGA toolchains, CUDA/OpenGL, and modern web stacks, and often prioritizes privacy-preserving, censorship-resistant architectures—an unusual combination in embedded vision startups.
8 years of coding experience
6 years of employment as a software developer