My name is Isabella and I'm a Computer Science major.
I just finished my second year at Vassar College and plan to graduate on May of 2017.
You can reach me at cubisa94 at gmail dot com
Dr.Jain
is an Assistant Professor at the University of Florida in the Department
of Computer and Information Science and Engineering. She is passionate about her research,
focusing on Human Centered computer graphics, or
the relationship between artistic visual content and computer technologies.
Before coming to University of Florida, Dr. Jain worked at Texas Instruments and as
adjunct faculty at University of Texas at Dallas.
You may reach her at ejain at cise dot ufl dot edu.
I am working on projects in which people will watch videos and answer questions based on their experience. I am also learning to collect behavioral data including pose data using the Kinect, heart rate data using a BITalino board, and Eye coordinate data using the RED eye scanner.
This week was about Logistics and set up. The first day I met with Dr.Jain and we talked about how I should
get set up in the lab and get registered with the University and department office this turned out to be harder than expected.
I toured the department and the lab
and got my own key so I could access the building. We also talked about the projects I will be working on. She asked me to meet
With another student who had previously been working on the project so that He could explain the procedures and the code.
I met with Beida, a fellow student and researcher in the environmental studies department whom I will be working with on this project,
and she piloted a trial of the experiment. It was really exciting to see how the experiment worked and what I would be dedicating my
time to.The rest of the week was Reading the documentation for this experiment and understanding the procedures.
I learned how to prepare the videos needed for the experiment.
For the video feedback project, I researched the best website
to use for crowd sourcing.
This week I continued to work on understanding the code and procedures that Beida had demonstrated.
I find this challenging because we have such different coding style, but it is really cool to
see how things could be done differently than how I would have done them. I also find myself
challenged because I have scarcely used matlab code before, and am having to learn that as I go.
This is nice, however, since I am a hands on learner.
I am also having a really fun time working on this website! It's cool to be learning so many new things
in such a short amount of time.
This week I did a lot of reading on what it means to be ethical during experiments and studies. This was for a training required for conducting studies. Although the material was really dry and presented in a way that was not at all engaging (through an online reading course), I made the best of it and learned why we need to teach scientist ethics, seeing the studies and experiments people have conducted in the past. I would have never thought of some of the precautionary thing we need to take into account.
I also worked with setting up the Kinect. And thanks to the documentation left for us by the students who wrote some of the code we are using, it was not as hard to tackle as it would have been just using the Kinect documentation (which I think is more than lacking).
This week focused on designing the studies. The team designed some questions that we would ask the subjects after they have been presented the visual stimuli for the video project. Then we split off and I worked with another student to implement the questions in the code.
My partner and I were assigned the task of finding and reading research papers that were relevant to the pose Kinect project and recorded our findings. This proved to be harder than anticipated. I thought that there would have been way more research papers published on the subject, and that they would have been a lot easier to find. I followed the citations on each of the papers and found that about 2 or 3 papers in, they would lead me in a circle. But In the end, we were able to accumulate a comprehensive list.
I also researched and compared 6 different commercial heart rate monitor devices based on their performance, sample rate, and development software.
This week I focused on debugging experiment code and storing the data we collect. Since last week we implemented all the right questions, now, we had to make sure that the answers were being recorded correctly and in a format that would be easy to read. I had a test on Friday to see if I had learned all the procedures and I was sorry to see the code gave us trouble. Although we had run all the code to make sure it was working beforehand, I had not practiced with all the equipment on and when we turned some of it on and tried to run the experiment, it crashed. This showed me the importance of trial runs that include all of the equipment.
I also worked with a partner on developing an app that would help us validate the data collected for the Kinect pose project (code name: playback app). She took the lead on that, and I focused on continuing to research the studies that have already been conducted and finding reasons why we should follow in their footsteps. I had no idea that this was part of the research process and was astounded at how draining it was. All in all, I was able to move forward in the work, and am super excited to hear that we will be conducting the study soon.
After this week, I can say that I have gained a really in depth understanding of the projects I am working on. After reading the research papers relevant to the Pose project, I was assigned the task of compiling a comprehensive list of previously studied body gestures. I made an excel spread sheet Which contained the gesture type, name, description and the papers that it appeared in. I found out the relevancy of each gesture to our project and why we should choose to study it again.
I also got to work hands on with the heart rate monitor and find out how it turns ECG data to heart rats in BPM. This was very interesting, yet frustrating, I feel as though I might have enjoyed it better if the circumstance of my research was different. But since some code which I had not written or modified seemed to just all of a sudden break, it was quite annoying. In the end, I was able to get the code up and running again. I learned the importance of saving prior work so that I could revert to previously working code.
This week was more coding intensive. I had to use the knowledge that I gained last week on how to measure heart rate data and apply it so that I could interpret the data correctly and write meaningful information to text files. I also had to test that the data was being written out correctly by comparing what I wrote to the file and the original .mat files.
We also met with the wildlife project group to discuss the progress that we had made. I shared my progress with the Heart rate and we decided that changes needed to be made to the experiment's code so that the data we collected would be more meaningful.
I also worked on a conversion rate from meters to pixels so that we could use it in the playback app, since kinect makes it really hard to access any information on the functions that they use. The 2 major problems being that a) the measurements were in meters so when drawing on the screen, everything was way too big, and b) we had to convert from the cartesian plane (the center origin used by kinect to store joint coordinates) to screen coordinates (which locate the axis on the top left of the screen with y increasing as you go further down the screen). I found this pretty cool since I had taken (and loved) linear algebra this past semester and as this was just a translation which could be defined as matrix multiplication. I had a really great time soling this!
I got the summer flu this week and was not able to do that much, but what I did do was really fun! I worked on a presentation about my REU experience which I am to present this coming Monday in front of other summer research students and their mentors.
I also worked on a proposal for a project which I am to apply the research of the heart rate devices I did, to complete. Unfortunately, we need to be granted special permission to access the heart rate monitor's developer kit. I sent them an email a while back asking if they would be able to give this to us, but they have not gotten back to me. I worry that this might delay the progress of the project.
The progress with the Wildlife project was really good this week! I was able to pilot the study with 2 fellow students who work at the lab.
This week we started the user studies for the wildlife project. I was really looking forward to having the first participant come in to see how it all went. To prepare, other than the piloting last week, I reviewed all the procedures in the user studies room to get comfortable with the space. I also got the payment gift cards from Dr. Jain as well as the consent form. I was not that nervous to actually conduct the study the first time seeing that the first participant was a student from the lab, so it was a familiar face. But I was more nervous that Dr. Jain and 2 other student were watching from behind the one way mirror.
It went really well, although we did encounter some problems. I was not clear as to where to put the heart rate monitor stickers and they put them on the wrong side of their body. We also ran over time by about 3 minutes. Since we pay our participants for their time, it is unethical to take time that is not paid for.
After the study I met with Dr. Jain and discussed what we could do to make the time shorter. We had talked with the study research team earlier that we could cut a video from the list if the study was taking too long. She told me to do so and inform the rest of the team that we were making this cut.
After cutting the video, We ran a second study to see how the time had changed. This time the participant was a student from the wildlife ecology department, so I wondered how it might go different. The time ran over again, this time by about 4 minutes. It seems that because the first participant was used to some of the procedures since they also at the human centered computing lab, they were more familiar with the process and therefore took less time.
I informed the team of this and we decided it was best to hold off on any further studies so we could get the time down closer to 30 minutes.
This week I worked on some data verification code for the data gathered on the wildlife project. I made a function which lets you choose 1 or 2 files depending on the data you have gathered and plots out the heart rate and gaze points to verify that they are within a reasonable range. For heart rate, this means that they are within 65 bpm and 110 bpm. For gaze points, this means that they are within the pixel dimensions of the screen, or at ( -99, -99) which means that the eyes were not recorded at that point (they were out of range from the eye tracker to capture, or they were looking off the screen) .
I was also introduced to another student, Melissa, who was to take on the project after I leave. We met a few times and I taught her the procedures of the study and how to use the experiments code, as well as the analysis code that I had written. Dr. Jain also asked us to work together to fix a few things on the code like to write some more data out to a file that was useful to the analysis of the study. We were able to do this and I feel that this gave Melissa a better understanding of the code and a certain level of comfort that I would have liked when I got there.
Teaching her how to do all of this also helped me a lot to better understand what I had been doing all summer. The questions that she asked were sometimes different than the ones I think I would have asked, so it was cool to see a different perspective and to work together to find the answer.
Dr. Anthony, one of the other professors in the human computer Interactions lab made brownies for the lab to say goodbye to the DREU students. It was fun to see each other and share our experiences and final thoughts on the program and what we experienced.