Cell Behavior Video Classification Challenge, a benchmark for computer vision methods in time-lapse microscopy
Abstract
The classification of microscopy videos capturing complex cellular behaviors is crucial for understanding and quantifying the dynamics of biological processes over time. However, it remains a frontier in computer vision, requiring approaches that effectively model the shape and motion of objects without rigid boundaries, extract hierarchical spatiotemporal features from entire image sequences rather than static frames, and account for multiple objects within the field of view. To this end, we organized the Cell Behavior Video Classification Challenge (CBVCC), benchmarking 35 methods based on three approaches: classification of tracking-derived features, end-to-end deep learning architectures to directly learn spatiotemporal features from the entire video sequence without explicit cell tracking, or ensembling tracking-derived with image-derived features. We discuss the results achieved by the participants and compare the potential and limitations of each approach, serving as a basis to foster the development of computer vision methods for studying cellular dynamics.
Cite
@article{arxiv.2601.10250,
title = {Cell Behavior Video Classification Challenge, a benchmark for computer vision methods in time-lapse microscopy},
author = {Raffaella Fiamma Cabini and Deborah Barkauskas and Guangyu Chen and Zhi-Qi Cheng and David E Cicchetti and Judith Drazba and Rodrigo Fernandez-Gonzalez and Raymond Hawkins and Yujia Hu and Jyoti Kini and Charles LeWarne and Xufeng Lin and Sai Preethi Nakkina and John W Peterson and Koert Schreurs and Ayushi Singh and Kumaran Bala Kandan Viswanathan and Inge MN Wortel and Sanjian Zhang and Rolf Krause and Santiago Fernandez Gonzalez and Diego Ulisse Pizzagalli},
journal= {arXiv preprint arXiv:2601.10250},
year = {2026}
}