Team Status Report
RISKS
Our greatest risk currently is related to a potential change in project Our proposed use of seam carving does not seem to have been attempted previously (at least in published research). We risk the possibility that it won’t work well for what we have imagined, which is smoothly shortening the length of a video. Because of this, our TA Mrunmayi has suggested that we develop a couple of alternate projects and decide on a final project next week. We are managing this risk with two efforts: We are brainstorming new project ideas related to our current one, and we are developing a more refined prototype of our project in software. The latter will help us to better understand whether it will be possible to smoothly shorten a video.
DESIGN CHANGES
Our main design decision this week was to decide on an FPGA board for our project. Our application is highly memory-constrained, and of the SoC boards available, only the Zynq-7045 had the desired on-chip block RAM. More details and specifications are detailed in Maxwell’s status report. This decision will inform much of our future work—the core of the SoC (Arm A9) determines our compute power, the amount of memory determines how our data must flow and where intermediate products can be stored, and the architecture of the board will inform how we connect to peripherals and interface between the fabric and the SoC.
UPDATED SCHEDULE
We initially intended to spend these two weeks on implementing the 2-D seam carving algorithm. We have decided to skip that part and start working on video carving right away as we realized the two have little correlation. The rest of our tentative schedule stays the same. More detailed schedule (a gantt chart) will be provided after we meet with our new faculty point of contact, Professor Low, to finalize our project.
Maxwell Johnson
Personal work:
I was the sole group member present at Monday’s class and one of two at Wednesday’s. As such, I’ve attempted to organize my group members and meet our goals. My main tasks in this capacity have been:
- Organizing a meeting
- I attempted to find time when all members of our team are free. It took a week for everybody to fill out a group calendar, but we have finally established the intersection of our availabilities, which we sent to Professor Low to organize a meeting with him.
- Updating absent team members
- After each class period, I sent an email with updates from that day’s class to my team.
- Assigning tasks to group members
- Included in my emails was an assignment of tasks for each group member. The intent was to allow the groupmates who missed class to continue working on our project.
- Requesting space for our team
- Requesting equipment (FPGA board) for our project
This was complicated by the fact that we changed sections to join the other group working on seam carving, then changed to work under the direction of Professor Low.
My main technical work was doing research and making a decision on the FPGA board we want to use. See a summary of results here: FPGA comparison.
I feel our progress is currently behind, due to the turmoil of the previous week and some uncertainty about whether our project will be viable. Our team status doc has some details on this uncertainty and our plans for dealing with it. I’ve met my schedule goals of deciding on a board for a project and beginning to learn the Vivado toolchain. The Windows machine at our workstation has Vivado 2018.1 installed. My goals for this week are to upload a test design to our FPGA and to begin learning the toolchain of interfacing with the SoC.
John Zhang
I will start this report by explicitly stating that I did not spend 12+ hours on our project due to personal health issues. I got food poisoning after a valentine’s day themed dinner this past Sunday, and I got very sick (103 fever). Since I had multiple things due at Tuesday midnight, a midterm at 8:30 am the next day, and another midterm on Thursday, I did not get to rest and recover from the sickness until Thursday afternoon. I took Friday off and that was when the fever completely stopped. Therefore, I wasn’t able to put in significant efforts throughout the week. I was, however, trying my best to stay in contact with my team and reading emails to stay updated on everything. Today, I set up Visual Studio in preparation to implementing the video carving algorithm.
For the following week, my main goal would be to implement the video carving algorithm. As of now, we do not have a concrete “right answer” to what the algorithm should look like, so I expect it to take much research, and experimentation.
This week has been rough, health wise. But I am ready to make up for it in the following weeks.
Riki Khorana