The datasets here are used to test 3DeeCellTracker, a deep-learning based software to track cells in 3D time lapse images. Data description: unet_training train_image: A 3D image for training the 3D U-Net. train_label: Cell/non-cell regions for train_image. valid_image: Another 3D image for validating the 3D U-Net. valid_label: Cell/non-cell regions for valid_iamge. single_mode_worm1 data: The 3D time lapse image of a semi-immobilized worm. manual_vol: The manually corrected segmentation in volume 1 models: The pre-trained 3D U-Net and FFN for this dataset ensemble_mode_worm4 data: The 3D time lapse image of a "straightened" freely moving worm. manual_vol: The manually corrected segmentation in volume 1 models: The pre-trained 3D U-Net and FFN for this dataset Links for further information: 1. GitHub repository of [3DeeCellTracker][1] 2. The eLife paper ([Wen el al. 2021][2]) describing the algorithm, the data and the tracking results 3. The files under "ensemble_mode_worm4/data" are a part of [another dataset][3] in a previous study ([Nguyen et al. PLOS Comput. Biol. 2017][4]) [1]: https://github.com/WenChentao/3DeeCellTracker [2]: https://elifesciences.org/articles/59187 [3]: https://ieee-dataport.org/open-access/tracking-neurons-moving-and-deforming-brain-dataset [4]: https://doi.org/10.1371/journal.pcbi.1005517