Week 1: May 26 - 30
I arrived at Clemson University's campus Sunday night excited about what the next 10 weeks would have to offer. The following day I finally had the opportunity to meet my mentor, Dr. Sophie Joerg and discuss the projects we will be working on for the summer. The main project involves investigating the importance of eye motion in virtual characters. This project seeks to model human eye motion on animated figure to improve its overall effectiveness. There is also a side project (Realism vs. Reaction) which intends to make animated characters in video games more realistic based on their reaction time. Dr. Joerg explained digital production arts and motion capture. I was quite intrigued by the way it worked and how it will be used in the projects. I have begun with some background research on the main project by reading different chapters and articles of professors that have done past research on the topic. I also started to get familiar with Autodesk Maya, which is a 3D animation software program that allows you to design animated figures. So far, I really enjoy using this program and learning all of its features. On Friday, my mentor and I created a 9-week outline of what should be accomplished throughout the summer. As of right now it seems like a lot of work, but I am looking forward to the challenge.
Week 2: June 2 - 6
By the start of the second week, I had settled in good and become comfortable with the project. I continued reading through the tutorials on Autodesk Maya to become better with 3D design. I was able to start building a face for our animation model. Tuesday I read and summarized a research paper given by my mentor that was similar to our main project. It contained key information about past experiments on the subject and methods for completing them. I found the paper to be very helpful because it gave me ideas on how I would like to conduct my research. My mentor also introduced a website called digitaltutors.com that assisted in designing the face model. I was able to find a lesson about ways to model eyes and animate them using Maya. I will admit, the process of designing in Maya can be a bit frustrating at times, but the result is quite rewarding. Later in the week I began to learn how the eye tracker works. The Dikablis eye tracker records eye movement and generates data based on change in the eye's angle. Unfortunately there was a problem with the D-Lab software used by the tracker, so I plan to explore other measures.
By the end of the week the eyes and face of the model were created, besides a few minor details. On Friday, I met with my mentor Dr. Joerg and discussed the articles I read and how I could use the info from them to implement details of the project. Lastly, I also posted flyers to recruit participants for our side project.
By the end of the week the eyes and face of the model were created, besides a few minor details. On Friday, I met with my mentor Dr. Joerg and discussed the articles I read and how I could use the info from them to implement details of the project. Lastly, I also posted flyers to recruit participants for our side project.
Week 3: June 9 - 13
In week 3, I continued to work on the 3D design in Maya. I attempted to fix the details to where the iris of the eyes looked more realistic. Before they looked very cartoonist, but I've given them a more sound look. In the previous week, I had issues with the Dikablis eye-tracker so my mentor and I revised our ten-week tentative schedule. Instead of working on data-driven animation (eye-tracker) first, we decided to do procedural animation. Procedural animation is produced through new or enhanced rules that are based off past research. The articles I have read in the first two weeks will help animate our model. Tuesday I read and took notes on another paper by Normoyole about perception of the human eye.
Mid-week I started learning about the Python programing language. Python code is written to animate our model. I found the language to be easy since it is similar to Java where I have experience in. For the rest of the week I worked on the design and Python commands. Within a day I was able to get the eye motion working. On Friday I worked on the eyelids. I was also able to complete an outline (algorithm) for the implementation of data-driven animation.
Mid-week I started learning about the Python programing language. Python code is written to animate our model. I found the language to be easy since it is similar to Java where I have experience in. For the rest of the week I worked on the design and Python commands. Within a day I was able to get the eye motion working. On Friday I worked on the eyelids. I was also able to complete an outline (algorithm) for the implementation of data-driven animation.
Week 4: June 16 - 20
At the start of week four I met with my mentor to discuss how the procedural animation method would work. I was unsure at first how to go about animating the model without data, but Dr. Joerg gave me some pointers. In Python Maya, I must code how the eye will move, and that involves direction, velocity and magnitude. Throughout the week I worked on the algorithm and code for the problem with the help of past research papers from other professors. So far two functions have been implemented and are working properly. The animation has limited eye motion and blinks twice.
This week was quite different compared to the first few. During our Monday morning meeting, Dr. Joerg and I also discussed how to conduct the experiment for the side project. Starting Wednesday, we tested different individuals on a video game to determine the effect of responsiveness and motion quality on a player's performance. Testing will continue next week.
This week was quite different compared to the first few. During our Monday morning meeting, Dr. Joerg and I also discussed how to conduct the experiment for the side project. Starting Wednesday, we tested different individuals on a video game to determine the effect of responsiveness and motion quality on a player's performance. Testing will continue next week.
Week 5: June 23 - 27
Week five mostly involved testing more people on the video game for our side project. Men and women were tested on 3 different conditions and prompted to answer questions on how they felt about their performance and game quality. I found it interesting how people would respond to the game. Many thought it was cool and fun. The different conditions really had an effect on the way people played. Testing was concluded at the end of the week with a total of 21 participants.
I worked on the main project of eye animation in segments this week. Right now I have the eye saccades working for about 8 seconds. Next week I hope to have the eyelid saccades working with the eyes simultaneously. The video game testing took more time than I thought, so I've fallen behind in the weekly schedule. My goal is to have the procedural animation working by next Tuesday, then begin using the eye-tracker.
I worked on the main project of eye animation in segments this week. Right now I have the eye saccades working for about 8 seconds. Next week I hope to have the eyelid saccades working with the eyes simultaneously. The video game testing took more time than I thought, so I've fallen behind in the weekly schedule. My goal is to have the procedural animation working by next Tuesday, then begin using the eye-tracker.
Week 6: June 30 - July 3
This week turned out to be fun and exciting. I started off the week continuing with the eyelid saccades and also quicker eye movements. It took a while to debug the program, but I was glad I managed to find the problem on my own, which makes a better programmer. Tuesday visitors from the company Ergoneers (manufacturers of the Dikablis Eye-tracker) came to give a presentation on their new software, and also fix our eye-tracker. I learned a few tricks about the software and how they could be used to obtain data for my animation. Unfortunately they weren't able to fix our eye-tracker, which puts a halt on the data-driven part of my project. The Ergoneers will be back next week to work on the tracker more. Wednesday I met with Dr. Joerg to discuss the progress of my animation and further steps to be taken. Since I got the eye & eyelid saccades working, the next step was to make the motions look more random. I worked on that for the rest of the week.
Week 7: July 7 - 11
At the start of the week I got the random eye blinks to work, however my model was blinking too often. After consulting with Dr. Joerg, the algorithm was changed to make the model's blinks look natural. I also altered the code to randomize saccade directions based on the data obtained from the Badler research paper read earlier in the summer. More functions were added to the code to make it easier to read.
As it turns out, the Dikablis Eye-tracker did not need any extra fixing, so I worked with the tracker starting mid-week. Dr. Duchowski gave a great demonstration on how to use the tracker along with the D-Lab and Vicon software. I have a great feel of how to operate the tracker. My next task is to find a subject to do the experiment, then import the data into Maya.
Next Tuesday I will be doing a presentation on my project to other DREU students and mentors here at Clemson. To prepare, I've been working on the presentation throughout the week along with my mentor. Also, Dr. Joerg added an extra condition for the experiment in our side project. Friday I tested 10 people on the condition.
As it turns out, the Dikablis Eye-tracker did not need any extra fixing, so I worked with the tracker starting mid-week. Dr. Duchowski gave a great demonstration on how to use the tracker along with the D-Lab and Vicon software. I have a great feel of how to operate the tracker. My next task is to find a subject to do the experiment, then import the data into Maya.
Next Tuesday I will be doing a presentation on my project to other DREU students and mentors here at Clemson. To prepare, I've been working on the presentation throughout the week along with my mentor. Also, Dr. Joerg added an extra condition for the experiment in our side project. Friday I tested 10 people on the condition.
Week 8: July 14 - 18
This week began in chaos. Monday I worked all day and late into the night on my final presentation. The presentation itself was completed earlier that afternoon, but I had issues adding my animated clip to it. In order to make the animation in Maya into an actual video, I needed to perform a playblast, or render out the frames. Unfortunately, neither of those methods worked so I wasn't able to add it to my presentation. However, despite the mishap I think my presentation went very well. I enjoyed learning about the other DREU student projects and what they accomplished this summer.
For the remainder of the week I worked on reestablishing my 3D model. In the long process of attempting to render out my animation, I lost my entire design, which was quite upsetting. By the end of the week my new model was complete, and I was able to do a playblast on it. Next, I need to conduct my perceptual experiment using the data-driven method. I still have yet to figure out how to save information from Dikablis & Vicon to a file. I plan to take on this task next week.
For the remainder of the week I worked on reestablishing my 3D model. In the long process of attempting to render out my animation, I lost my entire design, which was quite upsetting. By the end of the week my new model was complete, and I was able to do a playblast on it. Next, I need to conduct my perceptual experiment using the data-driven method. I still have yet to figure out how to save information from Dikablis & Vicon to a file. I plan to take on this task next week.
Week 9: July 21 - 25
The week began with some disappointing news. Monday I attempted to compile the necessary files in order to process data collected from Dikablis and Vicon. I received the files from a grad student who did a similar study about eye-tracking. Unfortunately, one of the files (vicon.h) was missing, so now we are trying to locate it. This puts data-driven animation on hold. Hopefully, we will have the files before my DREU is over. In the meantime, I altered the script for procedural animation to enhance its realistic features. I added a wait time between each eye rotation and also implemented saccade duration, which gives a varying time of how long it will take for each rotation to occur. The blink function is now working perfectly, with eye blinks happening on average 5 times in the animation. I believe my animation is very close to being natural-looking, but there is room for improvement.
Week 10: July 28 - August 1
Well, I've reached my tenth and final week in the DREU program. This week consisted of finalizing details in my project and completing the final report. Unfortunately, we were not able to animate our 3D model using the eye-tracker, but I have an idea of how it would've panned out. To substitute, I questioned five people about our procedural model, asking how realistic they believed the character was based on its eye motions. In this quick experiment, I learned for sure that the eyes, eyelid saccades and blinks are essential aspects for creating realistic eye movements in virtual characters.
This summer I learned many things that will assist me in my career as a computer scienctist. My programming and critical thinking skills are definitely better. Being at Clemson University for the summer was definitely an unforgettable experience. Because of my experiences with mentors, grad students, and fellow DREU students, I plan to pursue a masters in Computer Science. I want to thank my mentor, Sophie Joerg for being a great advisor, and the DREU program for granting me this opportunity. I'm really looking forward to next summer!
This summer I learned many things that will assist me in my career as a computer scienctist. My programming and critical thinking skills are definitely better. Being at Clemson University for the summer was definitely an unforgettable experience. Because of my experiences with mentors, grad students, and fellow DREU students, I plan to pursue a masters in Computer Science. I want to thank my mentor, Sophie Joerg for being a great advisor, and the DREU program for granting me this opportunity. I'm really looking forward to next summer!