KongNet is a multi-headed neural network using a shared encoder and specialized decoders for joint nuclei centroid, mask, and contour prediction, achieving first or second place in MONKEY and MIDOG challenges plus top-three in PUMA and SOTA on PanNuke and CoNIC.
Leveraging pathology foundation models for panoptic segmentation of melanoma in h&e images,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
KongNet: A Multi-headed Deep Learning Model for Detection and Classification of Nuclei in Histopathology Images
KongNet is a multi-headed neural network using a shared encoder and specialized decoders for joint nuclei centroid, mask, and contour prediction, achieving first or second place in MONKEY and MIDOG challenges plus top-three in PUMA and SOTA on PanNuke and CoNIC.