Fringe counting by machine learning algorithm!

We have mentioned previously that we are hoping to use volunteer classifications to train a machine learning algorithm to help us do further (and quicker!) analysis on the high speed electronic speckle pattern interferometry images like the ones we are looking at in this project.

Our collaborator at Belmont University in Tennessee has made quite a bit of progress using the averages of all the classifications done so far. Here is a short clip of some of his results that show how his algorithm has identified antinodes and counted the fringes of each of them:

Animation of algorithm identifying antinodes on steelpan imaged by ESPI

This doesn’t mean we don’t need you to keep working on our project!!  It is critical that we keep going in order to validate the algorithm and, if possible, make it better.

Thanks for your interest in the project – we will have more updates soon!


Recent progress!!

We recently aggregated all the recent classifications our volunteers have made this summer and compiled new animations showing the identified antinode regions.

Here is one of the animations:

Animation of steelpan strike and analysis
Animation of one of the steelpan strikes with volunteer classifications overlaid.

We have shown the above animation (an earlier version) previously – and here is another animation of a different strike on the drum:

Second animation of aggregated classifications.
Second animation of aggregated classifications.

The top animation is a collection of 740 frames having at least 5 classifications, while the lower animation is a collection of 655 frames having at least 5 classifications.

Newsletter #5

Hello Steelpan Vibrations Volunteers,

We are happy to announce to you a final wave of information for the summer.



Our data aggregation software is in its final stages of development. We have been going through a lot of error-reducing and debugging of our program. Check out the blog post with images from our program here



We successfully had our twitch live stream last Wednesday night. Professor Morrison went over various aspects of the project and we hope it clears up any questions many of you might have!  You can even check out the replay of the live stream here!



We are now a listed project on the SciStarter website! If you are a fan of citizen science projects, please visit our page there where you can bookmark our page and check yourself off as a volunteer for our project. 



As the summer is coming to an end, so is most of the team working on this project. Thursday, July 26 will be the last day the two undergraduates will be actively working on this project. That, however, does not mean this project is ending. We still are in need of two complete subject sets, so Professor Morrison can analyze them! Please continue to support this project as it will stay active while volunteers like you continue making classifications and bringing interesting discussions to the table.



We continue to thank each of you for participating in this project since the beginning of the summer.  We hope that our constant updates on the project have proven to entertain and inspire you to continue making those classifications. We also want to thank our moderators for doing such a good job at keeping things lively and posted on our project page. Without any of you, this project would not be what it is today. The team leaving hopes to see the volunteers continue to make classifications and make this project thrive! 


Thanks for all your help,
Andrew & the Steelpan Vibrations Team

Updated classification animation

Our volunteers have added THOUSANDS of classifications so far this summer – THANK YOU! Take a look at this animation which strings together many of the frames in sequence with the average ellipses indicated in blue:


This animation has roughly DOUBLE the amount of data that we had last time we shared a view like this on our blog. We’re getting a lot closer to having enough classifications to do a deeper analysis of the vibrations in the steelpan.

We are making great progresskeep going!!

Screenshots of our custom software for processing classifications

Throughout this summer, we have been working hard on developing software for us to aggregate all of the classifications our volunteers have made. We request a text file from Zooniverse which contains all of your wonderful classifications!

One member of the Steelpan Vibrations team has been developing a graphical user interface (GUI) for us to use to overlay average ellipse measurements on top of the original images from our project. Below is a screenshot of the GUI which shows the software displaying an unmodified image:


Our software sorts all of the classifications and is able to parse the data to recognize groups of similar ellipses which are marking the same antinode. Then the average ellipse is calculated for each antinode and displayed in the GUI:


You can take a look at our code here:

Newsletter #4

Here is the Newsletter sent out to the Zooniverse subscribers of Steelpan Vibrations:


Greetings Steelpan Vibrations Volunteers,
We are happy to give you another weekly update on the Steelpan Vibrations project!

Some exciting news: we are planning to do a live stream event this week on Wednesday night on! Read about it on our blog and sign up for the event on our Facebook page.

We are still in the process of creating a stable and user-friendly program that can interpret our classifications for analyzing and feeding into a machine-learning algorithm. Our GitHub link is still available on the front page of the project!

As suggested by a volunteer on our talk pages, we have updated our workflow to get rid of the first step of counting the number of antinode regions you see in an image. Instead, you are immediately asked to start drawing the ellipses around antinodes.

To all our new and long-time volunteers, we want to thank each one of you for providing classifications to our project. In the past week, you have contributed over 2500 classifications which is fantastic for our project! While the numbers have been dwindling down as expected, we hope that you continue participating throughout the summer and beyond. Just know that we couldn’t continue this research without you!

Next week is the last week that the summer research team will be actively working on this project. That does not mean this project is ending as it will still continue on until we reach our final goal of 200,000 classifications and retire all of our subject image sets!

Once again, we thank each and every one of the volunteers for providing us feedback and making classifications. We know summer can get quite busy with vacations and spending more time with family, but we do hope that every one of you continues to participate in Steelpan Vibrations into the Fall and Winter!

Thanks for all your help,
Andrew & the Steelpan Vibrations Team

Livestream on Twitch this Wednesday, July 18th @ 9PM CT

Hello Steelpan Vibrations Volunteers,

From the limited responses we have received on the best time to live stream and working around our schedule, we have decided to host a livestream of us performing classifications on Twitch this Wednesday, July 18th around 9 pm Central time! The team will be online, including Professor Morrison.

The purpose of this live stream is to show our volunteers how we perform a classification and in doing so will give volunteers like you a few tips on how to make a proper classification. We also hope to hold a live Q and A session to answer any questions you may have for us! It will be hosted on this channel:

We hope to see you there on the live stream!

-Steelpan Vibrations Team

Newsletter #3

Hello Steelpan Volunteers,

We thank you for your patience from the previous 3 weeks without any newletters, but we are back in action to keep you up to speed!


As you may have noticed, Zooniverse recently sent out a call to action on our behalf to the entire Zooniverse community. To sum it up: Our goals have temporarily narrowed and we are seeking many volunteers to help us reach that goal! Whether you are new to our project or have been with us from the start, we are thankful for all of you who are helping us out.


We have been working on our data processing code for future analysis and for feeding information into a machine-learning AI. The progress on that has been going well. We are publicly sharing our code on GitHub if you are curious on how it runs. The GitHub link is here and will be on our main page in Zooniverse:


We have also been updating our blog regularly. A recent blog post explains why we need a large number of classifications and gives a few examples of classifications with missed antinode regions:

Another post talks about how we are analyzing our data including classification examples and some critique:

This post delves into the origins of this Steelpan Vibrations project and how it came to be, check it out here:

We also visualized our goals by creating a hypothetical histogram based on how we want our progress to look. Read about it here:


That is all we have for you now. As always, we appreciate you helping us out with the Steelpan Vibrations project and we hope you continue to volunteer, especially with our new goal in mind!

– Andrew & the Steelpan Vibrations Team

Welcome New Volunteers and LiveStream Next Week!

Hello Steelpan Vibration Volunteers,

As always, we appreciate your time and dedication to this project. We also want to welcome the new volunteers to this project and thank you for your time and dedication on our project, there is lots of information to read on this blog with continuous updates until the end of the summer!

Just to inform the new volunteers, we are STILL a long way from our goal despite narrowing our goals since the beginning of summer. At this rate, we need about 2082 per day to reach our goal of classifying, so please keep it up and tell your friends and family about this project!

On another set of announcements, we want to ask all our Steelpan Vibrations volunteers AND followers if they would be interested in viewing a live stream (not longer than about 30 minutes) where the research team does classifications of their own and answers any questions that the volunteers want to pose. We believe this is critical since there may be questions you might or might not have had when making classifications and we want to make sure the entire classification process is clear to the steelpan vibrations volunteers!

We would like to know when the best time would be to stream for you. This might not serve everyone’s schedule, but we will be sure to try to accommodate as many people as possible! Let us know in this survey here:

As always, we thank you for following this project and keeping it lively. Without you, this project would not see any success!

-Steelpan Vibrations Team