News & EventsDepartment Events
Events
-
Mar29
EVENT DETAILS
As AI techniques continue to advance, the efficient deployment of deep neural networks on resource-constrained devices becomes increasingly appealing yet challenging. Simultaneously, the proliferation of powerful AI technologies has raised significant concerns about sustainability and fairness, demanding increased attention from the community. This talk presents two novel software-hardware co-designs for improving the efficiency and sustainability of deep learning models. The first part introduces a hardware-efficient adaptive token pruning framework for Vision Transformers (ViTs) on embedded FPGA, HeatViT, which achieves significant speedup under similar model accuracy compared to the state-of-the-art. HeatViT is the first end-to-end accelerator for ViT on embedded FPGA and also achieve practical speedup by data-level compression for the first time. The second presents PackQViT and Agile-Quant, a paradigm of the efficient implementation for transformer-based models by sub-8-bit packed quantization and SIMD-based optimization for computing kernels. Our framework can achieve better task performance than state-of-the-art ViTs and LLMs with significant acceleration and power saving on edge processors, such as mobile CPU, Raspberry Pi and RISC-V. This work not only marks the first successful implementation of the LLM on the edge but also addresses the previous limitation where edge processors struggled to efficiently handle sub-8-bit computations. At the conclusion of the presentation, the speaker will discuss today's challenges related to AI sustainability and fairness and outline her research plans aimed at addressing these issues.
TIME Friday, March 29, 2024 at 11:00 AM - 12:00 PM
LOCATION L440, Technological Institute map it
CONTACT Catherine Healey catherine.healey@northwestern.edu EMAIL
CALENDAR Department of Electrical and Computer Engineering (ECE)
-
Apr2
EVENT DETAILS
MemComputing is a new approach to computation that employs memory (time non-locality) to both process and store information on the same physical location. Memory is a remarkable feature since it can generate correlations between the machine units that do not decay exponentially, rather algebraically, both temporally and spatially. A memcomputing machine then navigates its phase space by following specific trajectories which showcase these long-range correlations, namely during dynamics the machine can change the values of just a few or as many variables in the problem specification as needed to reach the solution efficiently. Massimiliano will discuss the fundamental reasons behind this efficiency and show many applications of MemComputing to combinatorial optimization problems, machine learning, and quantum problems. He will finally discuss the path to its hardware realization. Work supported by DARPA, DOE, NSF, CMRR, and MemComputing, Inc.
TIME Tuesday, April 2, 2024 at 11:00 AM - 12:00 PM
LOCATION L440, Technological Institute map it
CONTACT Catherine Healey catherine.healey@northwestern.edu EMAIL
CALENDAR Department of Electrical and Computer Engineering (ECE)
-
Apr4
EVENT DETAILS
The number of autonomous agents in everyday life is expected to scale dramatically over the next decade driven by advances in machine learning, compute power and data availability. These cyber-physical systems will need edge perception: i.e. the capability of being situationally aware in diverse environments and the ability to make real-time decisions exploiting data collected using a network of multi-modal sensors.
One omnipresent information carrier for edge perception is the thermal signal emanating from all objects at non-zero temperature. Here, we build the information theoretic foundations of thermal perception to prove that infrared radiation, even in pitch darkness, carries equivalent amounts of information per detected photon as daytime visible radiation. This opens a new frontier of heat-assisted detection and ranging (HADAR [1]) which can have societal impact similar to existing techniques of Radar, Sonar and Lidar. I will discuss the design and deployment of the thermal voyager : an autonomous navigation agent which exploits passive infrared thermal imaging for semantic segmentation, 3D depth perception and path planning.
Finally, I will explain how this research presents disruptive opportunities for collaborations cross-cutting all the core areas within ECE: computer engineering, signals and systems as well as photonics/solid-state devices.
TIME Thursday, April 4, 2024 at 11:00 AM - 12:00 PM
LOCATION L440, Technological Institute map it
CONTACT Catherine Healey catherine.healey@northwestern.edu EMAIL
CALENDAR Department of Electrical and Computer Engineering (ECE)
-
Jun10
EVENT DETAILSmore info
McCormick School of Engineering PhD Hooding and Master’s Degree Recognition Ceremony
TIME Monday, June 10, 2024 at 9:00 AM - 11:00 AM
LOCATION Welsh-Ryan Arena
CONTACT Amy Pokrass amy.pokrass@northwestern.edu EMAIL
CALENDAR McCormick School of Engineering and Applied Science
-
Jun10
TIME Monday, June 10, 2024 at 2:00 PM - 4:00 PM
LOCATION Welsh-Ryan Arena
CONTACT Amy Pokrass amy.pokrass@northwestern.edu EMAIL
CALENDAR McCormick School of Engineering and Applied Science