Spatiotemporal (ST) data collected by sensors can be represented as multi-variate time series, which is a sequence of data points listed in an order of time. Despite the vast amount of useful information, the ST data usually suffer from the issue of missing or incomplete data, which also limits its applications. Imputation is one viable solution and is often used to prepossess the data for further applications. However, in practice, n practice, spatiotemporal data imputation is quite difficult due to the complexity of spatiotemporal dependencies with dynamic changes in the traffic network and is a crucial prepossessing task for further applications. Existing approaches mostly only capture the temporal dependencies in time series or static spatial dependencies. They fail to directly model the spatiotemporal dependencies, and the representation ability of the models is relatively limited.