All posts by ab127

Poster MANIA!

1 Comment

This week was hectic- I finalized my program on Wednesday with a lot of help from Yong, our mentor. Although, I wanted to the computer to calculate the heart-rate- I ended up settling with raw manual data analysis.

I was working on my poster like  a madman on Thursday and Friday. I am quite proud of my final product. I haven’t received my feedback, so that should be exciting. I used a QR code on my poster as well- thanks to Jimmy for that idea.

Bukhari1

I got my results and conclusion- I would have love to increase my sample size, but I ran out of time. As a group, we have to start planning how we will teach next week- I’m interested to see how high school students will be different compared to my middle-school students.

Picture of Shelia’s program- while I comparing our PPG outputs.
Pictures of my program

Stiff Upper Lip- week 3

4 Comments

I wanted to start off by mentioning that this has been a TOUGH week. I almost named my blog post this week ” Hard Knock Life!”, but I decided to focus on the positive aspect instead.

Monday and Tuesday

Crashing and burning, through the first two days. I had problems download two particular libraries- “matplotlib” and “scify”. I had to resort to using command prompt-I was very uncomfortable using it. After spending two days, just jumping from one video to another trying to find why my libraries were not working, I brought in my own laptop. No luck with that either.

Success!!!

Wednesday morning, I got some external help and finally got my libraries up and running in Sublime Text.  I spend the next two days trying to draw a rectangle, slicing the array from the rectangle to collect data for skin color changes in the forehead area.

How to add, average and collect data?

On Friday, I successfully got the data. Next, I need to check if I’m slicing it correctly. I need to store this data in a list. Then, graph it using matplotlib. I’ll be spending some time learning about matplotlib today. I’m somewhat content that I managed to collect data at least, considering how I started the week.

I’ll be attaching videos from my progress, even some that might be really simple- stuff I programmed in Week 1. Just to show how far I’ve come in my journey.

face_detection

color_tracking

finalcode_sofar

Trotting along- 2nd week!

Image Leave a comment

June 21, 2019

Azka Bukhari- Hogg Middle School: CS Educator

Face Detection

I started this week by trying to build a program that would detect a face in a still picture. After much frustration, Yong helped me add another line of code to my Haar Cascade. This helped my face detection program work- next step would be face detection via the webcam feed.

Facial Detection Code Through Webcam Feed

Yong and Presentation

Yong, our mentor, came by on the first day of the week. He laid out a step-by-step guide to help us emulate distancePPG. The first step was ‘Thresholding’- which means to turn an image into a 2D array-so it displays in only two colors: black and white. I had to put a hold on my facial detection plan and move to thresholding. Yong has been helping throughout this week- we started using Slack to share our code-he has helped me consistently when I get stuck.

Screenshot of Yong sharing his presentation on Slack.

Tuesday and Trotting

By Tuesday, I completed thresholding and facial detection with a still image and with the webcam feed. I took a break to focus on NumPy; I followed a tutorial online to help me learn a bit more about NumPy. This helped me learn the basics of NumPy. I knew I would be working from home on Wednesday- so I took all my notes home.

My notes on NumPy

Gifts on Friday

Allen brought us food and a Pulse Oximeter. I’ve been thinking my TeachEngineering and using an Adafruit Circuit Playground to build a visual thermometer. This might allow me to create a Maker Challenge and I would like to use the IB Design Cycle, which is very close to the Engineering Design Cycle on the website. It is still quite early, so I might change my mind. One of the obvious benefits is that my advanced students have been using the Adafruit Circuit Playground-so they will be comfortable in using it. I wrap-up this week  contently, I’ve completed thresholding, face detection and I’m almost done with shape detection of a face-but it might be more complicated than it looks.

Pulse Oximeter- how we will collect the ‘ground truth’ reading for the heart rate, we will be comparing this to our distancePPG reading to check for accuracy.

 

Steep Learning Curve

2 Comments

June 14, 2019

Azka Bukhari- Hogg Middle School: CS Educator

Prior Knowledge

The second week at Rice University’s PATHS-UP Research Experience for Teachers 2019-2020 has been quite eventful. I wanted to go ahead and summarize the extent of my research prior to starting the research. I googled the PATHS-UP RET page and found some interesting information. I tried to read the research on distancePPG; I understand that the whole concept of this project is to create a system of detecting vital signs in a non-invasive manner that is accessible to the under-served population. By reaching to teachers, the plan is to not only spread aware of STEM fields, but to educate, inform and encourage students about health in the under-served community.

Orientation & Yong’s Presentation

PATHS-UP team and Yong: our mentor

I presented my PPT early in the orientation and I met the rest of the team. Our mentor, Yong presented the research project that we would be working on- distancePPG.  PPG stands for photoplethysmography which is an optical method to measure cardiac-synchronous blood volume changes in face, finger and earlobe. In broad strokes, Yong explained the procedure of distancePPG and the various tools and libraries used by distance PPG and some of the challenges of using distancePPG. In layman’s term, distancePPG uses webcam monitoring while shining a light on a person- as the heart pumps bloods, the volume of blood in the arteries and capillaries change, this  leads to a small change in the skin color. DistancePPG measures this small change in skin color as the light shines on the person- some of the light reflect, some light is absorbed, the computer program calculates this light reflection.

Fig. 2

Fig. 2: Overall steps involved in distancePPG algorithm for estimating camera-based PPG.DistancePPG: Robust non-contact vital signs monitoring using a camera Biomed Opt Express. 2015 May 1;6(5):1565-1588.

Python is the programming language that will be used and two specific libraries that were mentioned were OpenCV and NumPy. OpenCV (Open Computer Vision) is a library of programming functions that focuses on real-life computer vision; it is FREE! This week I downloaded OpenCV and I have tried a YouTube tutorial (*tutorial link attached) – as basic introduction on how to use OpenCV with Python. NumPy, the other library mentioned, adds support for large, multi-dimensional arrays and matrices, along with high-level mathematical functions to open on these arrays. My experience with NumPy is minimal and I look forward to learning late next week. It was quite unclear how I am going to be doing my own research at the moment.

Scalable Health Labs & Reading

We visited the Scalable Health Lab to see the extent of their research. Mary Jin presented one of her projects that might potentially lead to detecting cancer cells without a biopsy. While, Anil Kumar Vadathya summarized his project as F.L.A.S.H- which stands for Family Level Assessment of on-Screen Time at Home. His project idea was to help reduce negative health impact of over-exposure to technology by monitoring the total number of hours children were watching TV. This was really interesting as this is quite close to our project when he broke down the process- the face detection and gaze detection.

As a group, we came together on Thursday and summarized what we read in the research paper. We went over the pseudo-code to understand how the program works. We weren’t sure about how well we did- but it was a starting point to tackle a huge task.

 

Learning Python

Christina introduced the team to edX, I started with the Absolute Beginners Course for learning Python. However, Pamela recommended codeHS.com, I picked this course because I can use this with my advanced classes and this is an excellent chance for me to experience it myself. This week, I have managed to get halfway through the Python course- I realized that I know a lot of fundamental concepts such as loops, conditionals and functions.

Screenshot of my CodeHS progress

Later in the week, Jimmy brought his own device that he created **. I have attached a picture of this experience- I will try this in my own classroom.  He decided to expose his students to a pulse oximeter that he purchased last year- this was followed by students creating their own pulse sensors using an Arduino and a pulse sensor bought from Amazon***.  Although, it would be replicating what Jimmy did, I think my middle school students would be blown away by the real-life application of Computer Science in the field of Health.

Walkaways

I am leaving this week- a lot more energized. I realized that being confused in part of the process of research. I am not used to feeling ‘lost’ and ‘confused’ being a teacher- this usually leads to PANIC! I have decided to use my confusion as a learning experience- when I am confused in Python, I realize that this is how some students might feel when they are learning code. When I listen to Yong’s presentation and Jimmy’s implementation, I learned that that it is my job as an educator to expose students to ‘very difficult’ concepts- things they might not grasp straight away. If we want our students to grow, we must give them challenging tasks. I have two ideas about where I would like to go with my own research project. First, I would like to create face detection and later- face recognition program. Jimmy has already done this, but I would like to try this myself. I can create a lesson with pseudo-code that I could use with my middle-schoolers. My second idea is quite interesting, but might not be do-able! I would like to see what would happen if the R.O.I was flashed with different colored lights- how would this affect the data being collected.

I’m still a little unsure about how my research project and poster will look like. I have seen some examples, but I am feeling a little anxious that I won’t do well. Shout out to Yong for breaking down tasks for us to do next week! Best thing this week! I really appreciate Professor Yong. d

*https://www.youtube.com/watch?v=YY9f-6u2Q_c

**https://learn.adafruit.com/make-it-pulse?view=all

***https://www.amazon.com/Asiawill-Pulsesensor-Sensor-Module-Arduino/dp/B00MB1LPUA?source=ps-sl-shoppingads-lpcontext&psc=1