Intelligent Identification and Prediction Mineral Resources Deposit Based on Deep Learning
Problem
Predicting ore deposit locations is hindered by scarce labeled geological images, small irregular features with high image similarity, and hard-to-quantify prospecting factors.
Why it matters
Deposit prediction guides costly mineral exploration decisions in the mining sector.
What the AI did
Built a multiscale feature attention framework (MFAF) on geoimage data, with an MFCA-Net module to mitigate scarce labeled images and an SE-Net channel attention module to quantify prospecting-factor influence.
Results
MFCA-Net reported to ease scarce-label problem 'to a certain extent'; no quantitative metrics stated.
Impact
Offers a deep learning approach to deposit identification despite scarce labeled geological images.
Record
| Venue | Sustainability |
|---|---|
| Published | 2023 |
| Work type | article |
| Authors | Le Gao, Kun Wang, Xin Zhang, Chen Wang |
| Citations | 15 (OpenAlex, at import time) |
| References | 57 |
| Model or method | MFAF (MFCA-Net, SE-Net modules) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/su151310269 |
| Open access | gold |
| Keywords | deep learning, mineral prediction, attention mechanism, geoimage, metallogenic prediction |
Importance rating
real but narrower advance
Deposit prediction guides costly mineral exploration decisions in the mining sector; a real but narrower methodological advance with no quantitative scale given.