A deep MIL model with attention, adaptive hard-sample weighting, and attention-driven hard-negative bag generation reports state-of-the-art classification on colon and breast cancer histopathology datasets.
Explaining t he stars: Weighted multiple-instance learning for aspect-based sentiment an alysis
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Deep Instance-Level Hard Negative Mining Model for Histopathology Images
A deep MIL model with attention, adaptive hard-sample weighting, and attention-driven hard-negative bag generation reports state-of-the-art classification on colon and breast cancer histopathology datasets.