12 million images later, Mars starts to make sense
12 million images later, Mars starts to make sense
https://news.asu.edu/20260430-science-and-technology-12-million-images-later-mars-starts-make-sense
Publish Date: 2026-04-30 17:00:00
Source Domain: news.asu.edu
- The vast amount of Mars images collected by orbiters has resulted in a challenge for scientists in understanding the planet due to the varied, incompatible formats and resolutions.
- Mirali Purohit, a doctoral student at Arizona State University’s School of Computing and Augmented Intelligence, aims to tackle this challenge by creating the Mars Orbital Model (MOMO), an AI model tailored specifically for Mars data.
- MOMO was developed by training on roughly 12 million Mars images, and it can identify small geological features as well as map large-scale geological activities.
- Despite common perceptions, Purohit highlights the diversity of Mars’s geological features due to Mars’ history and varied geological compositions.
- MOMO’s success allows for a holistic understanding of Mars, contributing to the study of its geological history and potentially revealing signs of past water or life.
- The goal for Purohit and her lab is to make MOMO, along with the training images, available to the scientific community to advance planetary science at an accelerated rate.
- Purohit also plans to merge orbital and rover imagery to give a comprehensive view of Mars.
- In the long-term, she aims to deploy models like MOMO to enable continuous data processing, improvement, and adaptation in real-world settings.
