FUN-ML: Fault Uncertainty for Machine Learning

Participation call

We are conducting a research study in the OLIVES lab at Georgia Tech for a seismic label uncertainty project. The goal of this research project is to generate a public dataset containing seismic fault annotations across varying levels of expertise, including novice, mid-level, and expert annotators, facilitated by the Amazon Mechanical Turk platform. 

Watch our instructional video for a quick walkthrough of the task. 

Eligibility criteria 

To participate in this study, you must: 

  • be between 18 – 89 years old, 
  • be able to consent without requiring surrogate consent from a legally authorized representative 
  • have access to a working computer (desktop or laptop) in which you can perform the task 

Compensation 

Upon completion of the full task, you will be rewarded with $71, with up to a $20 bonus based on an internal metric of annotation consistency. 

Participation 

If this opportunity sounds interesting to you and you meet all the criteria listed above, we would be delighted to have you participate in our study! If you wish to participate, just follow the steps described below.   

For any questions, please contact us at: jpacora3@gatech.edu and mohit.p@gatech.edu 

How to participate 

If you decide to be in this study, you will analyze a total of 480 images over a window of 8~9 hours (you may choose to do it in a single or multiple sessions). The entire image dataset is divided over 20 numbered batches on the platforms, which you may use to pace your progress. There is an instructional video in each task, which you only need to watch once. Make sure you watch the videos. There are four major phases for your participation.  

Phase 1: Watch the instructional video for a quick walkthrough of the task.  

Phase 2: Sign up for Mturk following the following steps.  

  1. Create an amazon account here. Make sure the account is US-based. If you already have an Amazon account, skip this step. 
  1. Log into the mturk worker sandbox here using your amazon account. 

Phase 3: Find task on worker sandbox.  

Once on the mturk worker sandbox, use the search bar on top to search for the requester username ‘olives.gatech’. This will display all available HITs for the task, numbered 1 to 20.  

Phase 4: Take the post-task survey  

Complete this 3-minute survey in which we will collect your mturk worker ID along with other information in order to properly process your payment.. 

Print Friendly, PDF & Email