CS698N: Recent Advances in Computer Vision

Gaurav Sharma, CSE, IITK

Computer Vision has rapidly progressed in recent years and is starting to appear in products used in
day to day life. The progress has been driven by both improvements in algorithms and computing
resources. Knowledge of such methods and technology is valued both in industy and academia.

In this course, we will look at topics in at least three exciting sub-areas of research in
Computer Vision among the following.

-- Human Analysis eg. pose estimation, facial analysis, attribute recognition, pedestrian detection
-- Language and Vision eg. image captioning, visual question answering
-- 3D computer vision eg. shape representation, reconstruction, affordance estimation
-- Image segmentation eg. semantic segmentation and multi resolution edge estimation, instance segmentation

The course would contain a significant project component addressing some of the main problems in
computer vision. The course would aim to equip the students with a moderately advanced understanding of
vision problems to prepare for a career oriented towards industrial or academic research.

The students wishing to participate in the course should have good foundations in basic maths and
probability. Good programming skills are also required for the project. Knowledge of machine
learning or willingness to learn the basics quickly are also needed.

The exact topics and lecture schedule will be decided later depending on the level of students
interested.