Using human-computation games to annotate accelerometer data
Jan 2018 – Present
Robust activity recognition algorithms depend on how well their training datasets have been prepared. An important part of preparing this training data set is to have high-quality annotations/labels on the entire dataset. However, if the dataset gets bigger and bigger, adding annotations to it becomes even more cumbersome for the research team. Often, these tasks are done manually on small datasets by researchers.
Therefore, in order to solve this computational problem, we have designed “Mobots” – a human-computation (i.e. crowdsourcing) game to annotate large accelerometer datasets (e.g., NHANES and UK BioBank datasets). In Mobots, players are shown snippets of accelerometer data that they match with a lab-based ground truth data (as shown below). This way, we can gather annotations on a very large dataset with the quality of lab-based ground truth. To access the current prototype of the game, please click here. We are carrying out several iterative tests on Mechanical Turk to improve this game.
Unity, AWS, C#, Actigraph, Actilife, R, PHP
Game Design, Research Design, Data Pre-processing, Level and Economy Design, Game Testing and Debugging
Game Design, Expert Interviews, Unity Programming, Database Design and Management, Data Analysis and Wrangling, Project Management
Prof. Seth Cooper (Asst. Prof, Northeastern University), Prof. Dinesh John (Asst. Prof., Northeastern University), Binod Thapa-Chhetry (Ph.D. Student, Northeastern University), Josh A Miller (Ph.D. Student, Northeastern University)
This is project is made possible by funding from the NIH BD2K (Big Data to Knowledge) project. Details soon.