Gondwana Geoscience is pleased to announce that teaming up with DeepSightX for the Gawler Challenge has resulted in a Merit Award for finishing in the Top 5 in the competition, out of a total of 59 submissions and more than 2000 registered participants. Only 1st and 2nd place getters received cash prizes, places 3 to 5 were unranked and received an Award of Merit. Positive feedback on the submission, entitled "Augmenting domain expert knowledge with AI | data-driven mineral & geological prediction" was received. Comments from the judges included, verbatim:
"Gravity classification algorithm is novel." "The most innovative modeling approach we've seen." "Domaining and geology classification is very clever."
The link to the submission, as well as others, is found here. https://unearthed.solutions/u/competitions/80/submissions
Results and reports from the Challenge, as well as the submission videos, can be found on the South Australian Geoscience Portal SARIG: GawlerChallengeResults
The gravity and domain classification work is based on image interpretation and classification work by Matthew Zengerer originally designed for working with Airborne Gravity Gradiometry data. Links to papers outlining this work can be found in the Conferences and Publications Menu on this website. At least two detailed papers outlining the main aspects of this work as well as the machine learning aspects are in either preparation or submission stages currently.
For more information regarding this work and how it can be applied to regional or detailed interpretative geological mapping as well as mineral targetting, please contact Gondwana Geoscience directly.