登录
  • #招生
  • #University of Michigan--Ann Arbor

Ph‌‌‍‍‌‍‍‌‍‌‍‍‌‍‍‍‍‍‍‌‍‌‍‍‍‍‍‌‍‌‌‌.D. openings at Michigan Tech for Intelligent/neuromorphic Robotic (Spring ...

MTU-ECE
729
3
Ph.D. positions with full RA/TA support are available in Hongyu An's research group.

Hongyu is an assistant professor at the Electrical and Computer Engineering Department of Michigan Tech. His research interest mainly focuses on Neuromorphic Computing, Neuromorphic Electronic Circuits Design using Emerging Devices, Spiking Neural Networks, Intelligent Robotics, and Medical Applications.

Neuromorphic computing (NC) is an emerging research topic that realizes artificial intelligence by mimicking and recreating the human brain, including neurons, synapses, Spiking Neural Networks. This approach involves both hardware design and algorithm development, such as circuit design, FPGA, and neural network algorithms, edge computing, etc. Recently, the group is applying NC on intelligent robotic and robotic learning. The project goal is to develop a self-learning robotic system that does not rely on massive datasets and can directly learn from constant interaction with surrounding environments.

The group closely collaborates with top research groups/institutes, including Intel neuromorphic research lab (Loihi chips), Sandia National Labs, Oak Ridge National Labs, institute of neuroinformatic of ETH Zurich, on robotics, spiking neural networks, and medical applications. The group also has multiple ClearPath huskies mobile robotics (Figure 1), emerging neuromorphic chips, Lidar, and other robotic sensors for our intelligent robot design.

Moreover, the group, as a neuromorphic community of intel labs, has the access to the Intel loihi neuromorphic chips, which are the top-tier neuromorphic computing platform in the market. the latest neuromorphic server named Pohoiki, which deployed in 2020. Pohoiki Springs is a large-scale system neuromorphic system built upon Intel Loihi neuromorphic chips. Each Pohoiki Spring consists of 24 Nahuku boards with 768 Loihi chips in total. The 768 chips contain a total of 98,304 neuromorphic cores incorporating a total of 100,663,296 neurons and 99,840,000,000 synapses.

Figure 1: Robot Platforms (ClearPath Jusky @ MTU)

The idea candidates are expected to have extensive experience on

; Understanding on robotic, such as ROS

; Understanding on edge computing

; Robotic sensors, e.g., Lidar, cameras, etc.

; Basic understanding of deep learning, neural networks

; Programming experience in one or more of the following languages: Python, C++, Java, etc.

; Programming experience on deep learning platforms: Pytorch, TensorFlow, etc.

; Ability to communicate effectively (both verbal and written)

If interested, please send me your CV, transcript(s), GRE, TOEFL (international applicants only), and sample publications if applicable to email address: hongyua@mtu.edu

Other detailed requirements from the department/college of Michigan Tech
3条回复
热度排序

发表回复