← Back to all news
arXiv AI By Yaroslav Prytula, Anton Popov, Dmytro Fishman

QCell: Recombining and Aligning Cell Queries for Overlapping Instance Segmentation

Read the original on arXiv AI →

QCell is a query‑based model designed to improve overlapping cell instance segmentation in microscopy images. It introduces an instance recombination module that decomposes and recombines query representations in latent space, allowing the model to reason about entire cell structures even when they overlap. Additionally, a contrastive query alignment objective is used to learn distinctive instance features and separate overlapping cell queries. The authors also present a new Organoid dataset benchmark and demonstrate that QCell surpasses state‑of‑the‑art methods, achieving +2.2 AP and +2.7 AJI on the ISBI2014 benchmark.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Sep 15

End-to-End Cell Detection via Instance-aware Graph Modeling

The paper introduces an end‑to‑end framework for detecting and classifying cells in pathology images by jointly modeling visual features and instance‑level interactions. It employs a dynamic graph construction module that builds cell graphs from learnable queries and an instance‑aware graph network that filters and reorganizes features, integrating appearance and relational evidence. Experiments on multiple staining protocols show the method surpasses existing approaches in both detection and classification accuracy.

By Ruochen Liu, Yalin Zheng, Jingxin Liu, Jianfeng Zhang, Shoujun Huang, Dexing Kong, Haofeng Li, Wei Lou
More like this →