My name is Chang Gao (高唱). I joined INI in January 2017 and have been working with Profs. Shih-Chii Liu and Tobi Delbruck on the Neuromorphic Processor Project (NPP). I am working on designing efficient hardware architectures to accelerate inference and training of recurrent neural networks (RNN) with the goal of realizing power-efficient real-time inference and incremental learning of large RNNs on mobile devices.
My CV: https://github.com/gaochangw/CV/blob/master/Chang_Gao_CV.pdf
During my PhD I have been actively working on:
- DeltaRNN accelerator (FPGA 2018, state-of-the-art real-time performance of RNN accelerators at the time)
- Speech recognition using DeltaRNN and Cochlear AMS1cV4 (ISCAS 2019)
- EdgeDRNN accelerator (AICAS 2020, Best Paper Award)
- RNN in real-time control systems (ICRA 2020)