Month: April 2019

Ajay

This week I spent mostly cleaning up the code base and verifying all the EC2 Connections were still working properly. We actually faced an issue with IT where they kept marking my raspberry pi as an insecure device and restricting network access. After repeated attempts to let them know that i had secured my device, eventually they realized that they had a bug in the vendor software they used to scan the network and our access was restored. The other half of the week I spent building the final presentation and practicing presenting. I also spent time verifying out final metrics and our mAP calculations. Most of these lined up close to our numbers that we wanted except that our matching characteristic had a little bit of a low percentage on accuracy. In general, the product is working pretty well but there are places where we could improve on it.

 

 

 

Team Report

Our in lab demo went well, we are still coordinating to practice our final demo and what it will look like. We are going to get started on the final report and poster next week.

 

 

 

 

 

Vayum

 

This week I created an API to integrate the data from Ajay’s facial recognition algorithm and seamlessly put it as a part of our web application. I used Django web requests, Postman, and wrote a stand alone program that looks for POST requests, parses the JSON data, and calculates a moving average of a continuous data stream. This part is to calculate the average time to wait based on the data we get from our cameras and raspberry pis. Overall, we still need to focus on figuring out how we want to integrate the data with Peters part along with occupancy part and the team. Other than that, I would say that we are up to date and on track.

Ajay:

This week I got most of the functionality working for the entire REID system. I finished the raspberry PI code so that now we can take photos and upload them to s3 buckets and call the appropriate POST endpoints to trigger storage/detection functions.   I actually found a 4x speedup within the raspberry pi code revolving around having the camera which was turning on and off. Instead I forced it to stay on for the duration of the program and as a result, we were able to speed up the overall operation of the raspberry pi by 4x. I finished writing the DB on the EC2 instance so now we can store our feature vectors and persist them in storage. The histogram functions are also finished so we can now extract them from the bounding box and persist them. I wrote the matching function as well which uses Bhatcharraya distance. This seems like it works well but we are still doing final testing to verify all the numbers. This following week I need to do the MAP calculations and do the other metric verification.

Team Report:

We are still on track to have a final demo but need to see how we are doing with the occupancy part of the project and focus the next week on testing and debugging that to see how accurate we are.

Week 7 Status Report

Vayum:
This week I focused on the heat map development portion of the project. This part turned out to be more challenging than I expected both with the actual design of the heat map and changing what we have to dynamic data that is always moving. Originally, I had a status of the just static data. We originally moved from one side to another by putting in fake data, but there was no way to really create a dynamic heat map of a table with the actual floor plan according to online resources. After doing a bunch of research, I had to import a bunch of modules that made incorporating these modules very difficult.

Overall, I would say I am a little behind in terms of this and need to focus especially on accelerating the process. I dont want to spend too much time being stuck on this and I would say that after this I would need t move on to other parts of the project.

Ajay:

Last week I wrote a small characteristic matching function to look at the histograms. This was a more brute force approach as I wanted to see how useful a brute force simple algorithm would work. It did not work that well so I will implement the Bhattacharya distance metric this week. We also got our raspberry pi’s this week so we will port our camera code over to the PI to ensure it works with that.

 

Team report:

Overall our team is slightly behind as we just received our raspberry pi’s this week. This is accounted in for our slack and we will be able to catch up