CS239: ML-driven Video Analytics Systems, Fall 2020

Instructor: Ravi Netravali
TA: Murali Ramanujam
Lectures: Monday/Wednesday 8am-9:50am
Office Hours

  • Ravi: by appointment (ravi@cs.ucla.edu)

  • Murali: by appointment (muralisr@g.ucla.edu)

Course Overview

Video cameras are pervasive. As camera deployments expand, organizations increasingly rely on analyzing video feeds to guide numerous applications including traffic monitoring, surveillance, and amber alert response. Key to the success of such applications has been recent advances in computer vision, particularly neural network (NN)-based techniques for highly accurate object detection and recognition. Though effective at answering high-level queries about video content, these NN-based pipelines are resource intensive in terms of network and server-side compute overheads. This class will explore a wide range of systems and machine learning optimizations to improve the efficiency of modern video analytics pipelines, without violating latency and query accuracy expectations.

Remote Lectures

Lectures will take place during the scheduled time (i.e., 8am-9:50am PT) on this Zoom link. The passcode for the Zoom meeting will be emailed to all enrolled students prior to the first lecture; if you are not yet enrolled but are interested in attending lecture, please email the staff for the passcode. Students are expected to attend lectures, actively participate, and follow video conferencing etiquette. Lectures will be recorded and the videos will be posted to CCLE.

We will, of course, accommodate time zone challenges. For those who are unable to attend all of the lectures due to time zone issues, the first thing to do is to inform the staff ASAP. Since participation is a major part of the course grade, the best way to earn participation points is by sending emails to the staff prior to lectures with 1) questions, and more importantly, 2) ideas that may prompt good discussion, e.g., extensions to the system, questioning system decisions, etc.


  • 40% Participation in paper discussions

  • 10% Paper summaries

  • 20% Paper presentation

  • 30% Final project (report and presentation)

Paper Summaries

This course will be entirely based on research papers. Prior to each class, students will be expected to read the listed research paper(s) and write up a brief summary for each. Paper summaries should be short and include the following components:

  • Paper description (few sentences) including what problem the paper tackles, and how it does so

  • Potential limitations of the solution (e.g., are any assumptions too general? are there scenarios where the proposed solution does not work?)

  • Potential extensions (e.g., ways to support more scenarios, other domains/applications where the proposed solution could be useful)

  • Any questions about the paper

Paper summaries should be submitted using this form, and are due by 10pm the night before each class. Students may skip paper summaries for up to 2 papers without any penalty. Also, students presenting a paper need not submit a summary.

Paper Presentations

For each paper, two students (or more, depending on enrollment) will be expected to present the paper and lead the discussion for it. Presentations are 30 minutes, should be “conference style”, and describe the domain and relevant background for the paper, the problem statement and challenges, the solution, results, and potential limitations and improvements. Non-presenters are expected to actively participate in the post-presentation discussions. Presenters are expected to come prepared with discussion points and non-presenters should come with ideas or questions for discussions (based on their paper summaries). Active participation will lead to a lively discussion that will benefit everyone.

Since this quarter's course will be remote, presenters should record their paper presentation video and send it to Murali (the TA) by 5pm the day before the corresponding lecture. Videos should be in the mp4 format and can be made, e.g., using Zoom; please contact the staff with any questions about making a video. Lecture will interleave watching the video, Q&A, and discussion.

Research Project

In addition to paper reading, this course will also include a quarter-long research project. Students will carry out projects individually. The goal with this research project is not necessarily to implement a research idea, but instead is to work entirely on formulating and motivating a potential research idea. In other words, your goal by the end of the quarter should be to convince the class that the idea you are interested in is 1) worth exploring, 2) challenging, and 3) previously unsolved. The scope of acceptable topics is quite large — anything related to any part of an end-to-end video analytics system (live or retrospective). Research directions should be ambitious, at the level of a conference paper (note that the course focuses only on motivation, not implementation, of a conference-level idea). It is encouraged to begin thinking about project topics early on in the quarter by reviewing the reading list/topics, and discussing with the instructor. The deliverables for the project are:

  • Meet with the instructors to discuss potential project directions (in class on Wednesday, November 18)

  • Project Idea presentations (10 minute in-class presentations on December 2 and December 7)

  • Final report (3 pages) detailing the problem, challenges, motivation results, related work, and potential solution direction in the Usenix conference paper format (due Wednesday, December 16 by 10pm).