Feature map coding method and device

The invention provides a feature map coding method and device. Relates to the technical field of data coding/decoding based on artificial intelligence (AI), in particular to the technical field of data coding/decoding based on a neural network. The feature map coding method comprises the following steps: acquiring a feature map comprising a plurality of feature elements; obtaining a probability distribution index value corresponding to each feature element, wherein the probability distribution index value is used for indicating probability distribution from a probability distribution set; each probability distribution in the probability distribution set corresponds to a probability distribution parameter, and the probability distribution parameter is obtained by calculating a probability distribution index value based on a learned nonlinear mapping model; and performing entropy coding on each feature element according to the probability distribution corresponding to each feature element to obtain a code stream corresponding to the feature graph. Through the method, the probability distribution number in the probability distribution set can be reduced while the performance is ensured.