[Defense] Anomaly Detection from Videos using Graph Convolution Networks
Friday, May 7, 2021
12:00 pm - 1:00 pm
In
Partial
Fulfillment
of
the
Requirements
for
the
Master
of
Science
Shoumik
Sharar
Chowdhury
will
defend
his
thesis
Anomaly
Detection
from
Videos
using
Graph
Convolution
Networks
Abstract
Hundreds of thousands of hours of video are recorded by surveillance cameras every day. Although much object detection and person detection and even anomaly detection is carried out on these video feeds~\cite{Kim2011}, the methods used have still been fairly traditional and repetitive: from bottom-up approaches using low-level features or leveraging the advent of convolutional neural networks or, more recently, image transformers. We propose a novel semi-supervised learning method to detect anomalies from a pedestrian dataset by representing each frame in the video feed as a graph. We create a graph embedding from video frames, where objects are treated as nodes and hand-crafted features between the objects are treated as edges. This embedding is then combined with convolutional features to detect anomalies.
Friday,
May
7,
2021
12:00PM
-
1:00PM
CT
Online
via
MS
Teams
Dr. Shishir Shah, thesis advisor
Faculty, students and the general public are invited.
