The data for our paper, "A high-precision method of segmenting complex postures in C. elegans and deep phenotyping to analyze lifespan", includes the following three components: Synthetic image dataset, CSB-1 dataset, and MD dataset, which originate from the paper "WormSwin: Instance segmentation of C. elegans using Vision Transformer". BBC010 dataset, sourced from the paper "Annotated high-throughput microscopy image sets for validation". Training weights for the Synthetic image dataset, CSB-1 dataset, MD dataset, and BBC010 dataset, as well as the pretrained weights used for worm tracking. The training weights from the synthetic image dataset can serve as pretrained weights for training on other datasets. Our experimental results are based on the average of multiple training runs; here, we have only uploaded one set of weights per dataset to facilitate reproducibility for readers. For more detailed information on the datasets, please refer to the relevant papers.