Congratulations to our labmates: Susan K. Meerdink, James Bocinsky, Alina Zare, Nick Kroeger, Connor H. McCurley, Daniel Shats and Paul D. Gader! Their paper, “Multi-Target Multiple Instance Learning for Hyperspectral Target Detection” was recently accepted to IEEE Transactions on Geoscience and Remote Sensing (TGRS).
In their paper, the authors introduce an approach to estimate multiple target signatures from training samples with imprecise labels. Experiments proved that the proposed method is capable of learning effective signatures for target detection in both synthetic and hyperspectral data.
Check out the paper here!