Job descriptions:

We are CV accelerator team under DSP department and we port Classical(non-neural-network) CV algorithm to our CVFlow engine(*).

In this job, you will need to be familiar with our CVFlow engine architecture for both NVP and GVP, and program them efficiently to run any non-NN CV algorithm requested.

The non-NN CV algorithm could be classical computer vision such as image processing filters,  optical flow and radar/lidar point-cloud processing.

During the algorithm porting, you will also need to optimize it as much as possible from both internal computation and memory I/O efficiency, based on the CVFlow architecture. 

*CVflow AI neural vector processor (NVP) engine enables high performance, low latency, and low-power NN processing. The NVP is enhanced to efficiently enable the latest advancements in NN inference, including support for transformer networks. Additionally, the CVflow’s general vector processor (GVP) provides traditional computer vision processing.

Job requirements:

- Experienced in ARM-based embedded system

- Strong C programming debugging skills

- Familiar with multithreaded programming in Linux or RTOS

- Familiar with image processing will be a plus

- Familiar with computer vision algorithms will be a plus