Microelectronics and Integrated-Systems with Neuro-centric Devices (MIND) Lab
MIND Lab provides an excellent research environment to the students. The students will work in the team of graduate, and undergraduate students, and PI.
Neuromorphic Computing and Artificial Intelligence
Artificial Intelligence (AI) has caught significant research attention. However, AI algorithms are compute resource intensive. Therefore, much emphasis has been devoted on developing specialized neuromorphic System on Chip (SoC) for accelerating AI recently. Some examples of these neuromorphic-SoC include Google’s TPU, Intel’s Nirvana and Loihi, and IBM’s TrueNorth. In spite of these significant advances, the performance of these processors are limited by the lack of novel reconfigurable memory devices that can mimic the adaptive characteristics of biological synapses. MIND lab researchers have invented new type of memory devices, referred to as, Low-Energy Evolutionary Channel Electronic Synapses (LEECHES) that can provide lucrative options for neuromorphic SoC. The student working on this project will work on this cutting-edge technology and develop neuromorphic SoC utilizing these novel memory devices. The student will perform the following tasks:
- Testing and Modeling of LEECHES in Verilog/VHDL
- Modeling and simulation of neural circuit modules around LEECHES
- Implementation of learning algorithms and performance evaluation of this novel neuromorphic SoC
- Publication and presentations.
- Training data collection on drones fitted with cameras and multiple sensors.
Trust and Security Issues in CMOS Active Pixel Sensors
CMOS Active Pixel Sensors (APS) serves the backbone of all digital cameras and video systems. Most of the autonomous systems heavily rely on data from cameras for making decisions. However, APS are prone to being tempered and attacked during operation. The student working on this project will evaluate various possible attacks on CMOS APS and remedies. The student will work on:
- Simulation of CMOS APS in HSPICE using 45nm transistor nodes.
- Identifying various techniques of introducing Trojans in APS.
- Understanding the impact of these Trojans on images generated from APS.
- Impact of this tempering on the behavior of autonomous systems.
- Identifying ways to mitigate it.
- Publication and presentations.
Rashmi Jha
Professor, CEAS - Elec Eng & Computer Science
385 Engineering Research Cntr
513-556-1361