We will demonstrate the following three technical achievements of our joint project:
1. Deployment of HarDNet on GPU (power consumption: > 200 Watts)
2. Deployment of HarDNet on FPGA (power consumption: several tens of Watts) [winning 2nd place in the FPGA track, LPCVC 2020]
3. Deployment of HarDNet on lightweight edge devices such as Raspberry Pi (power consumption: single-digit, < 10 Watts) [winning 3rd place in the DSP track and 4th place in the CPU track, LPCVC 2020]

Being performed on various computing platforms such as GPU, FPGA and AI edge device, HarDNet can consistently achieve highly competitive performance in terms of speed and accuracy. Especially, for the application of real-time semantic segmentation, HarDNet is ranked first around the world and has been recognized as "state of the art" (SOTA). Not only have we already deployed HarDNet on various platforms with different power budgets, but also we have been applying HarDNet and its variants on a variety of computer vision tasks besides those already done, including our LPCVC’20 winning projects.

1. The silicon valley smart voice chip manufacturer adopts the RNN acceleration solution developed by the team. Its high-end AI voice chip has been rolled off the line in 2020. The estimated output value of the chip is worth more than 100 million U.S. dollars.
2. Taiwan’s major memory chip manufacturer and our team work together on AI computing in memory technology to create a new blue ocean for next-generation chips.
3. Establish a startup company to provide momentum for the industry and cultivate AI talents for the country.

線上展網址：
https://tievirtual.twtm.com.tw/iframe/c56f1869-3239-453d-ab18-90ad1f1004c6?group=23bfb1fa-dd5b-4836-81a1-4a1809b1bae5&lang=en