skip to main content

Title: ADN: Artifact Disentanglement Network for Unsupervised Metal Artifact Reduction
Authors:
; ; ;
Award ID(s):
1722847
Publication Date:
NSF-PAR ID:
10168537
Journal Name:
IEEE Transactions on Medical Imaging
Volume:
39
Issue:
3
Page Range or eLocation-ID:
634 to 643
ISSN:
0278-0062
Sponsoring Org:
National Science Foundation
More Like this
  1. Being able to assess and calculate risks can positively impact an agent’s chances of survival. When other intelligent agents alter environments to create traps, the ability to detect such intended traps (and avoid them) could be life-saving. We investigate whether there are cases for which an agent’s ability to perceive intention through the assessment of environmental artifacts provides a measurable survival advantage. Our agents are virtual gophers assessing a series of room-like environments, which are potentially dangerous traps intended to harm them. Using statistical hypothesis tests based on configuration coherence, the gophers differentiate between designed traps and configurations that are randomly generated and most likely safe, allowing them access to the food contained within them. We find that gophers possessing the ability to perceive intention have significantly better survival outcomes than those without intention perception in most of the cases evaluated.
  2. As the making phenomenon becomes more prevalent, diverse, and vast, it becomes increasingly challenging to identify general temporal or spatial trends in types of making endeavors. Identifying trends in what participants are making is important to makerspace leaders who seek to understand the impact of the making phenomenon on the world or who are interested in broadening participation within their own maker contexts. This paper shows how topic modeling by means of LDA can be used to analyze maker artifacts, and illustrates how these types of insights can be used to make inferences about the making phenomenon, as well as to inform efforts to broaden participation.