April
15, 2025
Algorithms to
Implement Diversity Coding for Link Failures
in Networks
5:30 PM - 6:30 PM PT
Speaker:
Dr. Subarna Tripathi of Intel Labs -
Visual Algorithm Research group
Register
at: https://events.vtools.ieee.org/m/478569
In
this talk, exciting advancements of Dr.
Tripathi's research group, Visual
Algorithm Research, at Intel Labs will
be shared. In the first part of the
talk, several video representations
methods such as sparse graphs and sparse
transformers that are equipped with
long-form reasoning capability will be
covered. In the second part, the focus
is on the group's research addressing
egocentric use cases and several
associated algorithmic advancements. The
talk will be concluded with Dr.
Tripathi's team recent work around video
large language models.
About
the speaker:
Dr.
Subarna Tripathi is a research scientist at
Intel Labs, working in computer vision and
machine learning. She leads a team of
talented researchers working on long-term
video understanding & generation,
multimodal and structured representation
learning. As a co-chair of AI strategic
research sector, she helps oversee Intel's
global academic investment in AI. She is
serving as one of Intel's Center Lead
Liaison (CLL) for a JUMP2.0 center, CoCoSys
and Artificial Intelligence Hardware (AIHW)
Technical Advisory Board Member.
She
received her PhD in Electrical and Computer
Engineering from University of California
San Diego with Professor Truong Nguyen and
Professor Serge Belongie as her PhD
advisors. She is an alumna of Video
processing group at UC San Diego and SE(3)
computer vision group at Cornell Tech. She
has been an area chair of WiML since 2017
and a reviewer of CVPR, ECCV, ICCV, NeurIPS,
WACV, AAAI, ICLR, IEEE journals. Before
joining Intel, she worked in
STMicroelectronics in its Advanced System
Technology (AST) group for 6.5 years on
computer vision and video processing
domains. Prior to that she worked in Interra
Systems on video analyzer.
Dr. Tripathi received her MS Research from
Indian Institute of Technology, Delhi and a
Bachelor of Technology (Computer Science and
Engineering) from Kalyani Govt. Engineering
College. She is from Kalyani, a beautiful
town in West Bengal, India.
April
24, 2025
Volunteers
Needed: Annual Industry & Education
Summit - Student Project Showcase
Register
at: https://forms.gle/qPc1fYbFyP1dvRkFA
You are invited to the
Annual Industry + Education Student
Showcase, featuring career technical
education high school students from
throughout the county presenting projects
from innovation + design, manufacturing,
fabrication, and more. They are looking for
industry partners who would like to see what
students have been learning and making and
at the same time be part of a student's
journey through the discovery process.
The event is on April 24, 2025, 9:15 am -
1:30 pm PDT. However, the day is broken up
into 1 and 2-hour increments. There are
three ways to participate: Judge, Guest
Speaker, or Table Facilitator at lunch. The
time slots are below.
Rotation 1
Judge:
9:15 am - 11:00 am
Rotation 1 Guest
Speaker:
9:15 am - 11:00 am
Rotation 2
Judge:
10:15 am - 11:45 am
Networking Lunch Table
Facilitator: 11:15 am - 12:45 pm
April
29, 2025
AI for Research &
Development (AI4R&D): Revolutionizing
Innovation and Value Creation in
R&D
4:00 PM - 7:00 PM PT
Speaker:
Has Patel, Infologic
Register
at: https://events.vtools.ieee.org/m/477340
Integrating AI
into R&D processes presents a
promising future, offering significant
innovation and value-creation
opportunities across industries.
McKinsey estimates generative AI could
add $2.6 to $4.4 trillion in annual
value, with R&D and Product
Innovation use cases among the highest
contributors.
This talk will delve into the
practicality of the AI4R&D model,
addressing challenges in the R&D
lifecycle, such as information
overload, experiment design, AI risk
management, and research
commercialization “Valley of Death.”
Following the widely accepted and
federally developed Technology
Readiness Levels (TRLs) methodology,
strategies for integrating AI
throughout the
research-to-commercialization
lifecycle will be discussed to
increase efficiency and value
creation.
Attendees will learn how AI can
streamline processes, enhance
decision-making, and accelerate
innovation. The presentation will also
introduce AI-focused workforce
development, the future of AI, and
strategies for developing and
implementing AI-powered projects.
About
the speaker:
Mr. Has (Hasmukh) Patel
is passionate about research and development
(R&D) and innovation management. He has
gained extensive experience across various
sectors, including roles at Bell
Laboratories, GlaxoSmithKline
Pharmaceuticals, and B.P. Chemicals before
founding Infologic. Over fifteen years at
Infologic, he supported U.S. Defense
RDT&E organizations and developed
several innovation management models,
presenting them at defense and corporate
events. He is researching the AI-powered
Innovation and Value Creation in R&D
(AI4R&D™) model.
As a committed advocate, Mr. Patel
volunteered for the U.S. Department of
Commerce's NIST Generative A.I. Public
Working Group (NIST GAI-PWG). He represented
Infologic at the Department of Defense's
Manufacturing Readiness Levels (MRLs)
working group, the UCLA-managed Advanced
Manufacturing USA Institute—CESMII, and the
USC-led Advanced Manufacturing Partnership
for Southern California (AMP SoCal). He also
judges various business planning
competitions and the A.I. Innovation
Challenge, highlighting his dedication to
fostering innovation.
Presented by: IEEE Orange County
Computer Society, co-sponsored by
Buenaventura Computer Society
May
6, 2025
Algorithms
to Implement Diversity Coding for Link
Failures in Networks
7:00 PM - 8:00 PM PT
Speaker:
Ender Ayanoglu, Professor, UCI
Register at: https://events.vtools.ieee.org/m/466460
Diversity coding is a
form of network coding for link failure
recovery in communication networks. Since it
employs coding, there is no feedback
signaling, and that feature makes it very
fast. One approach, Diversity Coding Tree,
employs mixed integer programming and
results in very fast restoration. Another
approach is called Coded Path Protection and
employs integer linear programming and has
the advantage of small extra capacity. This
latter technique is based on former work
that considers a communication network as
consisting of bidirectional links. This
technique employs a mixed integer linear
programming formulation and results in
restoration times as fast as Diversity
Coding Tree with reduced extra capacity.
June
3, 2025
Machine
Learning in NextG Networks via
Generative Adversarial Networks
7:00 PM - 8:00 PM PT
Speaker:
Ender Ayanoglu, Professor, UCI
Register at: https://events.vtools.ieee.org/m/466461
Generative Adversarial
Networks (GANs) implement Machine Learning
(ML) algorithms that can address
competitive resource allocation problems
together with detection and mitigation of
anomalous behavior. In this talk, we
discuss their use in next-generation
(NextG) communications within the context
of cognitive networks to address i)
spectrum sharing, ii) detecting anomalies,
and iii) mitigating security attacks. GANs
have the following advantages. First, they
can learn and synthesize field data, which
can be costly, time consuming, and
non-repeatable. Second, they enable
pre-training classifiers by using
semi-supervised data. Third, they
facilitate increased resolution. Fourth,
they enable recovering corrupted bits in
the spectrum. The talk will provide basics
of GANs, a comparative discussion on
different kinds of GANs, performance
measures for GANs in computer vision and
image processing as well as wireless
applications, several datasets for
wireless applications, performance
measures for general classifiers, a survey
of the literature on GANs for i)–iii)
above, some simulation results, and future
research directions. In the spectrum
sharing problem, connections to cognitive
wireless networks are established.
Simulation results show that a particular
GAN implementation is better than a
convolutional auto encoder for an outlier
detection problem in spectrum sensing.
|