CryoAnomaly: Few-Shot Cryo-EM Particle Picking via Anomaly-Guided Hard Negative Suppression
Abstract
Cryo-electron microscopy (cryo-EM) is crucial for analyzing 3D biological structures, in which automated particle picking is essential for the workflow. However, fully supervised methods require extensive manual annotations. While few-shot learning offers a potential solution, existing approaches struggle to handle the diverse contaminations inherent in real micrographs owing to insufficient negative supervision, resulting in false positives that degrade the quality of the 3D reconstruction. Although synthetic data provides abundant and perfect labels, their use has primarily been restricted to validating identical proteins or augmenting full-shot training, leaving the potential for few-shot adaptation to novel proteins unexplored. In this study, we investigate the effective utilization of synthetic data for few-shot particle picking. We identify that direct transfer fails due to a ``clean-vs-contaminated'' Sim2Real gap. To overcome this, we propose CryoAnomaly, which is a framework that turns this gap into an advantage. By employing an anomaly detector trained on clean synthetic data, we identify real-world contaminants as anomalies and suppress them via a novel anomaly-guided hard negative suppression loss. On the CryoPPP benchmark, CryoAnomaly achieves the best picking accuracy and reconstruction resolution among state-of-the-art methods in the few-shot setting. The proposed anomaly-guided loss is confirmed to be effective on datasets with diverse contamination, where reliable pseudo-anomaly masks can be generated. Our code, dataset, and project page are available at: https://github.com/riku359/CryoAnomaly, https://huggingface.co/datasets/rikrikrik/CryoAnomaly, https://riku359.github.io/CryoAnomaly-page/.
Versions
Related
- LTK-YOLO: Lightweight Keypoint Detection for Robotic Tea Bud Picking
- Post-Ejection Reorientation of Rice Pot Seedlings for Automated Transplanting: Mechanism and Optimization of a Cam-Driven Continuous Picking System
- An inter-row navigation system design for a safflower picking robot
- Lightweight YOLOv8n marigold corolla detection and picking-point localization method based on ECR progressive channel compression and dual attention mechanisms
- A Robotic Picking Framework for Autonomous Singulation of Sweetpotato Slips into Chainpots.
BibTeX
@misc{itsuji2026cryoanomaly,
title = {{CryoAnomaly: Few-Shot Cryo-EM Particle Picking via Anomaly-Guided Hard Negative Suppression}},
author = {Riku Itsuji and Rintaro Otsubo and Ryo Fujii and Xingjian Li and Xiaolong Wu and Hideo Saito and Min Xu},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2608.28817},
eprint = {2608.28817},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.28817},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - CryoAnomaly: Few-Shot Cryo-EM Particle Picking via Anomaly-Guided Hard Negative Suppression
AU - Riku Itsuji
AU - Rintaro Otsubo
AU - Ryo Fujii
AU - Xingjian Li
AU - Xiaolong Wu
AU - Hideo Saito
AU - Min Xu
PY - 2026
DA - 2026/08/28
PB - arXiv
DO - 10.48550/arxiv.2608.28817
UR - https://arxiv.org/abs/2608.28817
AB - Cryo-electron microscopy (cryo-EM) is crucial for analyzing 3D biological structures, in which automated particle picking is essential for the workflow. However, fully supervised methods require extensive manual annotations. While few-shot learning offers a potential solution, existing approaches struggle to handle the diverse contaminations inherent in real micrographs owing to insufficient negative supervision, resulting in false positives that degrade the quality of the 3D reconstruction. Although synthetic data provides abundant and perfect labels, their use has primarily been restricted to validating identical proteins or augmenting full-shot training, leaving the potential for few-shot adaptation to novel proteins unexplored. In this study, we investigate the effective utilization of synthetic data for few-shot particle picking. We identify that direct transfer fails due to a ``clean-vs-contaminated'' Sim2Real gap. To overcome this, we propose CryoAnomaly, which is a framework that turns this gap into an advantage. By employing an anomaly detector trained on clean synthetic data, we identify real-world contaminants as anomalies and suppress them via a novel anomaly-guided hard negative suppression loss. On the CryoPPP benchmark, CryoAnomaly achieves the best picking accuracy and reconstruction resolution among state-of-the-art methods in the few-shot setting. The proposed anomaly-guided loss is confirmed to be effective on datasets with diverse contamination, where reliable pseudo-anomaly masks can be generated. Our code, dataset, and project page are available at: https://github.com/riku359/CryoAnomaly, https://huggingface.co/datasets/rikrikrik/CryoAnomaly, https://riku359.github.io/CryoAnomaly-page/.
ER -