Text recognition model training method, model training device and electronic equipment

Abstract

The invention discloses a text recognition model training method which comprises the steps that sample enhancement is carried out on a training text to obtain a plurality of training corpora, the training corpora are provided with labels, and the training text comprises a natural language text; inputting the training corpus into the first model to obtain a vectorized representation of the training corpus; calculating the weight information of each word in the single training corpus relative to the training corpus and the similarity information between the vectorized representations of the plurality of training corpuses; training the first model by using the similarity information and the weight information; inputting the vectorized representation into a second model to obtain a category score of the training corpus; and training a third model by using the category score and the vectorization representation, and determining the trained third model as a text recognition model. The invention further discloses a model training device, electronic equipment, a storage medium and a computer program product.

Description

technical field [0001] The invention belongs to the technical field of deep learning and small samples, and in particular relates to a text recognition model training method, a model training device, an electronic device, a storage medium and a computer program product. Background technique [0002] Deep learning models have achieved state-of-the-art results in tasks such as image classification and text classification. But the success of deep learning models relies heavily on large amounts of training data. In real-world scenarios, some categories have only a small amount of data or a small amount of labeled data, and labeling unlabeled data will consume a lot of time and manpower. In contrast, humans can learn quickly with only a small amount of data. For example, humans can quickly make judgments on newly given samples based on the previously learned knowledge combined with the labels of several currently given samples. Similar small-sample learning goals are to make ma...

AI Technical Summary

Problems solved by technology

Most existing methods only focus on the relationship between labeled and unlabeled, and do not fully supplement the category knowledge between samples. In addition, the existing few-shot learning methods do not consider the importance of different words in sentences to sentences.

Images