Model training method and device, pedestrian analysis method and device, equipment and storage medium

The invention provides a model training method and device, a pedestrian analysis method and device, equipment and a storage medium, and relates to the technical field of data processing. According to the model training method, network parameters in each level of network are updated layer by layer, so that better gradient updating of the parameters can be realized, and a better training effect is obtained. And the weight is updated while the parameter is updated in the training process, so that compared with a mode of manually setting the weight by virtue of experience, the scheme of the invention can obtain better weight based on parameter updating in the training process, and is beneficial to improving the training precision of the model. 本申请提供一种模型训练方法、行人分析方法、装置、设备及存储介质，涉及数据处理技术领域。其模型训练方法通过对每级网络中的网络参数进行逐层更新，从而可对参数实现更好的梯度更新，获得更好的训练效果。并且在训练过程中更新参数的同时更新权重，相比于人工凭借经验进行权重设置的方式，本申请的方案能在训练过程中基于参数的更新获得更好的权重，有利于提高模型的训练精度。