preview

Team Based Project Analysis

Decent Essays

The course, CS 160 Software Engineering, intends to educate on applying the fundamentals practice of software engineering through a team based project. For the term project, the requirement expected the group to develop a web application, which processes a user uploaded video to perform Facial Detection and display it as a playback video. Initially, the project description made it seem impossible achievement; however, the instructor’s introduction to software stack, which included OpenFace, FFmpeg, RDBMS systems, etc., allowed the team to construct a pipeline in order to deliver the project requirements. The initial step of our pipeline consists of accepting the user uploaded video, and serve it as an input to FFmpeg command, which would …show more content…

Initially, my idea was to FFmpeg’s node library to perform the tasks mentioned above; however, this approach turned out to be very challenging. Thus, I put forth the idea to use PHP to take advantage of shell execution in PHP. And eventually, my back-end role was set up shell executions in PHP, and then to retrieve output pass it along the pipeline. Besides my tasks, every member associated themselves with specific tasks to complete the remaining parts of the project. Mohnish took over the responsibility of configuring the setup of libraries and git repository of our project. Due to his knowledge of command line arguments, and working with MAMP, he was able to assist the team to install and build OpenFace, dlib, and FFmpeg libraries on our machines. Furthermore, he took responsibility of using OpenFace libraries to retrieve facial landmark points and track the head position of every image. While, Mohnish was working with OpenFace, Kirtan took over the responsibility of creating the schema on Firebase. Since, Kirtan had little exposure with Firebase API, he selected to implement the communication to database to either store or retrieve data, and also, assisted me to implement user authentication. Additionally, Kirtan figured out the implementation of tracking the pupil using eyelike. In amidst of constant code refactoring, Mohammad took the

Get Access