A transfer learning method and device

Abstract

The embodiment of the invention provides a transfer learning method and device, and the method comprises the steps: obtaining a preprocessed sub-sequence and a dense prediction network model, whereinthe preprocessed sub-sequence comprises a source domain sub-sequence and a target domain sub-sequence; Inputting the source domain sub-sequence and the target domain sub-sequence into a dense prediction network model in batches to obtain global shift loss; Obtaining a source domain label and a target domain pseudo label, and performing corresponding category centroid alignment according to the source domain label and the target domain pseudo label on the basis of a cosine distance loss and moving average method to obtain corresponding category shift loss; And obtaining total loss according tothe global shift loss and the corresponding category shift loss, and updating the dense prediction network model according to the total loss. According to the migration learning method and device provided by the embodiment of the invention, a multi-level unsupervised domain adaptation method is provided, alignment of edge distribution and condition distribution is completed, migration of a time series data intensive prediction model is realized, and the performance is more superior.

Description

technical field [0001] The present invention relates to the field of neural networks, in particular to a transfer learning method and device. Background technique [0002] With the rapid development and widespread popularization of wearable and portable sensor devices, human behavior recognition based on sensor data has been widely used in behavior monitoring, health monitoring, smart home, sleep monitoring and other fields, and has a good application prospect. At present, deep learning methods are widely used in human behavior recognition based on sensor data. This method extracts abstract pattern information from time series data generated by sensors, thereby achieving high-accuracy human behavior recognition. Dense prediction is currently the best performance human behavior recognition method. Considering that the duration of each human behavior is different and the exact boundary of the behavior is difficult to define, it is difficult to determine the optimal sliding win...

AI Technical Summary

Problems solved by technology

[0005] However, there is still a gap in the transfer learning method of dense prediction applied to human behavior recognition. The existing transfer learning method does not fully consider the unique deep neural network structure of the dense prediction model and its fine-grained recognition of each time point data. , does not transfer well for dense forecasting on time series data

Images