Lightweight fine-grained image recognition method for cross-layer feature interaction in weak supervision scene

Abstract

The invention discloses a lightweight fine-grained image recognition method for cross-layer feature interaction in a weak supervision scene, and the method comprises the steps: constructing a novel residual module through employing multi-layer aggregation grouping convolution to replace conventional convolution, and enabling the novel residual module to be directly embedded into a deep residual network frame, thereby achieving the lightweight of a basic network; then, performing modeling on the interaction between the features by calculating efficient low-rank approximate polynomial kernel pooling, compressing the feature description vector dimension, reducing the storage occupation and calculation cost of a classification full-connection layer, meanwhile, the pooling scheme enables the linear classifier to have the discrimination capability equivalent to that of a high-order polynomial kernel classifier, and the recognition precision is remarkably improved; and finally, using a cross-layer feature interaction network framework to combine the feature diversity, the feature learning and expression ability is enhanced, and the overfitting risk is reduced. The comprehensive performance of the lightweight fine-grained image recognition method based on cross-layer feature interaction in the weak supervision scene in the three aspects of recognition accuracy, calculation complexity and technical feasibility is at the current leading level.

Description

technical field [0001] The invention belongs to the field of computer vision, especially a method for fine-grained image recognition using image-level label weak supervision information combined with low-rank approximate polynomial kernel pooling and cross-layer feature interaction network framework, especially involving cross- A Lightweight Fine-grained Image Recognition Approach for Layer-Feature Interaction. Background technique [0002] With the rapid development of Internet technology, human society has entered the information age, and the total amount of data resources stored in various ways such as text, image, voice and video in the network is growing exponentially. Among them, image data has gradually become the mainstream information carrier because it is vivid and intuitive, and it is not restricted by region and language. It has broad application prospects and practical research significance. At the same time, the introduction of parallel computing theory and th...

AI Technical Summary

Problems solved by technology

Considering the phenomenon of information loss in the forward propagation process of the convolutional neural network, Bilinear CNN and various variant algorithms use the top-level convolutional activation of the deep neural network for bilinear pooling, but the features from a single convolutional layer It is not enough to describe the semantics of all key regions of the image, and directly treating it as a reference feature may lead to the loss of discriminative information that is important for fine-grained image recognition

In addition, bilinear pooling uses the matrix outer product operation to capture the pairwise correlation between feature channels, which significantly improves the recognition accuracy. However, this operation increases the dimension of the feature description vector

Images