Or Not Symbolism
I forgot to post this image yesterday. We were having an argument. Not about being racist, sexist, homophobic, or transphobic, but about the sizes of the milkshakes and whether or not a large is really worth the extra dollar. My argument was that the top (which makes the difference between the regular and large sizes) is basically just whipped cream and a cherry. So, it is not worth paying a dollar for just a cherry and some whipped cream in taller cup (less than an inch taller). Let me know if you agree or disagree.
On a related note, I am getting comfortable in Ames. I didn’t give enough credit to the town and I’m starting to realize that there’s more to it than meets the eye. I have developed a bit of a routine which might get disrupted as I start staying longer due to an increase in research duties.
A Portrait of Gabriel Kulp
Do you think this image is real? Hah, this image was CGI generated!
What about this one:
Wrong again! This image was also CGI generated. The point here being you should never believe you eyes and that hyper-realism is the death of trust in your vision.
Okay, hyper-realism is actually a bit of scary issue. Imagine you are watching the news and they use false imagery calling it live feed. For instance, let’s say CNN had a fake reporter live on a dangerous scene, but in reality they lied and used green screen to place their reporter in the scene. If you thought this example was to specific to not be true, you were right:
So even though I was joking about the topic above, I do believe this is an issue that should be discussed more. Let me know your thoughts on hyperreality.
“We experience something called “the death of the real”: we live our lives in the realm of hyperreality, connecting more and more deeply to things like television sitcoms, music videos, virtual reality games, or Disneyland, things that merely simulate reality.” – Doug Mann on Jean Baudrillard
I’m learning how to use ffmpeg mainly to convert videos into images. The image frames will be used in our dataset that trains the machine learning algorithm. (P.S. the GIF above is converting video to audio).
Hopefully, our algorithm will be able to annotate video data just as a human would. The video data for the TIMELI group consists of Traffic Incident Managers (TIMs) performing their usual work activity. The previous method for behavioral coding by user researchers included a manual process for logging different behaviors (the TIM is filing a report, the TIM is communicating to police by phone, etc). We hope to automate this process with a successful algorithm that does the same thing, but with more data. This will aid the user research process significantly allowing for more efficient interface research/design.
Quote of the Day: “The GIF is like our mind trying to understand everything” – Leilanie Morales
FFMPEG and JPEG,
Yesterday I hurt my leg,
An algorithm without rhythm,
The promise I renege,
A lit review is near,
*Please comment the end of this poem*
“Always design a thing by considering it in its next larger context – a chair in a room, a room in a house, a house in an environment, an environment in a city plan.” – Eliel Saarinen
An important aspect of HCI is the idea of user-centered design. This includes gaining an understanding of users’ backgrounds. TIMELI researchers want to revolutionize the traffic incident managers (TIMs) do their work, but in order to do that we had to put ourselves their shoes (the teams from previous years had already done this, but it was important for us as well).
What TIMs Do
This past Wednesday we had the opportunity to visit Iowa DoT’s Traffic Management Center (TMC). Although we weren’t conducting formal contextual interviews, we did observe and ask TIMs questions that would aid us in understanding their work. We gathered a lot of information through this process, but a main takeaway was redundancy in certain features. For instance, maps. Also, Jamiahus (Graduate Mentor), who has been working this project in previous years, noticed some changes with the interfaces. The main change was a new dashboard for TIMs that removed some issues with reporting incidents.
Purpose of Trip
As mentioned earlier, it’s important for user-interface researchers to examine the background of their users. This goes beyond the software requirements and looks at specific features that would improve the users interaction with the system. Overall, this was a very fun and informative trip. I’m glad to have some real world experience with HCI/UX research.
Quick Update: I’ve made my first machine learning model. This follows the Kaggle Cats vs. Dogs challenge. The image above shows that the overall accuracy of the algorithm is around 67% (ability to detect the object is a dog or a cat). This is just the beginning of TIMELI’s machine learning process; hopefully by the end of the program our algorithm will be able to detect and classify features of behavioral encoding data via videos of Traffic Incident Managers (TIMs).
What’s up next for TIMELI REUs?
We need to turn videos into frames of images for our algorithm to use as a dataset. We’ll be using ffmpeg (which I don’t have experience in) and I’m excited to learn how it works as well advancing our algorithm. Goodbye for now!
Today, I went down memory lane and looked at old code I wrote in python for my Intro to CS class. My team will be using python for our project and I wanted a bit of refresher. I looked at snippets of my code that I think could be useful for computer vision and machine learning problems. An interesting file I found was a graphics file that was provided by John Zelle. I never really looked too deep into the file, but after reviewing it further it provides a lot of information on programming with graphics in mind and would be a good practical resource. I don’t have much experience with machine learning, so I’m interested to learn more (I have a tiny bit of experience with neural networks, but not much further). I will be using Anaconda for this project because it has many libraries that are useful for machine learning.
I forgot to post about this earlier, but yesterday we got to repair some bikes and bring some back with us. We also received bike locks to use, so transportation will be a lot better around campus. That’s all to report so far, goodbye!
A task for our introductory lesson in HCI was to find two bad interfaces (virtual or physical):
The above image is Iowa State University’s ‘Join’ page. I wanted to show that usability issues are tied with accessibility issues. To demonstrate this, I used the WAVE web accessibility tool. The tool showed that the page had several issues like not having certain labels that are used by people with visual disabilities (text-to-speech tools).
I have a desk lamp similar to this and my issue is with the knob up top. Instead of a similar on/off button, I must turn the knob to the correct direction to turn the light on. This was very annoying early on because I would always turn it to the wrong direction. The device required a user to rely on the working memory (it is always better to have users rely on recognition rather recall). A simple and better design would have been a button on the bottom that when pressed turns on and when pressed again turns off.
I witnessed my second hail storm in Iowa a mere week later than the first one. Weather here is just as unpredictable as in Minnesota, but much more hail. My umbrella wasn’t as effective as I hoped it would be and I also fell in the mud (cartoon style). See below for a reenactment:
Others’ reaction while watching me (I’m looking at you Leilanie, Ohana, Victoria, and Teandre):
That is all for today, I still need to recover (mainly my dignity). Goodbye!
Communication is key. This has been a recurring piece of advice when working in teams and generally throughout life. I believe TIMELI’s (the group I’m a part of) really understood this in our meeting this past Wednesday. Although we had various interests, we worked through this by communicating and trying to find a middle ground between our interests and relevant work needed for TIMELI research.
Here’s some more fun things:
Quote of the Day
“Things in life never come easy, you always have to fight for them” – Leila Morales
Here’s a quick poem to wrap up the work day
Another day of C++,
I’m excited and sleepy,
Writing Code, I know I must,
So I can’t be weepy,
A problem I must solve,
*If you are reading this please finish the poem in the comments*
Wow, it is already week 2! I want to take some time to look back at the first couple of days.
I knew coming in the first week of the program would be difficult; I would be coming to a new environment (new state, new school, new people) and bringing my old along with me. By bringing the old with me, I mean homework and finals. I study at Carleton College which runs on a trimester schedule which means I am supposed to finish school in early June. I flew into Iowa for the VRAC REU on May 29 with a week of finals ahead of me.
Still Pretty Early On, But Later
Yesterday was my last day of finals and I had finished everything on time. But, this had some repercussions on my experience in Iowa. I am still pretty unfamiliar with the area because I haven’t had the chance to explore. The only places I had gone were the dorm, gym, and VRAC facility. We did go on a tour of the campus, but that isn’t how I usually get accustomed to places; it usually takes some time and exploration (getting lost).
I’m starting to get used to my surroundings! I went to Ames Summerfest and walmart which isn’t much for most people, but coming from small towns (Portland, ME and Northfield, MN), I was pleasantly surprised. I got some groceries and effectively used the transportation system (which is free!). I’ve also started to game a bit more (my roommates have open doors for me in terms of gaming possibilities) and am excited and worried to see where this goes.
As for my research, we will be heading down to meet some Traffic Incident Managers (TIM) in Des Moines. Hopefully, they will provide information on their experience that is beyond what I’ve read in previous research.