Improved method for learning discriminative segments in fine-grained identification

Abstract

The invention discloses an improved method for learning discriminative segments in fine-grained identification. The method comprises the following steps of: extracting segments with discrimination properties in an original image: obtaining a feature map by the original image through a convolutional pooling layer in a convolutional neural network, considering a vector of each space fixed position as a detector of a corresponding position segment in the original image, assuming that a detector, a discriminative region of which has highest response, in the original image is learnt, carrying out convolutional operation on the detector and the feature map so as to obtain a new response map, and selecting a position with a maximum value in the new response map so as to obtain the segments with discriminative properties; and learning features of the segments with the discriminative properties and carrying out classification by using the features: obtaining a local saliency map according to the segments with the discriminative properties, and encoding the local saliency map by using a space weighted Fisher vector. The method is capable of learning discriminative features more suitable forfine-grained identification tasks, and decreasing the interferences of background information in the discriminative segments so as to improve the classification precision.

Description

technical field [0001] The invention relates to discriminative block learning in fine-grained recognition. In particular, it relates to an improved method for discriminative patch learning in fine-grained recognition by spatially weighting image descriptors based on response maps to obtain spatially weighted Fisher vectors. Background technique [0002] In recent years, fine-grained recognition has attracted more and more attention in the field of target recognition. It is to identify subcategories of a large category of targets, such as flowers, birds, dogs, cars, etc., usually they have The same structure, so how to learn the discriminative features in the image becomes the main task of fine-grained recognition. [0003] In past research, the field of fine-grained recognition mainly includes two tasks: local localization and feature representation. In addition to image category labels, fine-grained data sets usually provide additional target bounding boxes and local part...

AI Technical Summary

Problems solved by technology

In addition to image category labels, fine-grained data sets usually provide additional target bounding boxes and local part annotations. Many previous works have more or less relied on these additional annotations, but fine-grained classification usually requires expert-level knowledge, and it is difficult for ordinary people to complete this task, which makes the cost of manual labeling very expensive

Images