Unifying Deep Local and Global Features for Image Search
This work addresses image search problems for computer vision applications, representing an incremental improvement by combining existing feature types.
The authors tackled image retrieval by unifying global and local features into a single deep model called DELG, achieving state-of-the-art results on datasets like Revisited Oxford and Paris and Google Landmarks v2.
Image retrieval is the problem of searching an image database for items that are similar to a query image. To address this task, two main types of image representations have been studied: global and local image features. In this work, our key contribution is to unify global and local features into a single deep model, enabling accurate retrieval with efficient feature extraction. We refer to the new model as DELG, standing for DEep Local and Global features. We leverage lessons from recent feature learning work and propose a model that combines generalized mean pooling for global features and attentive selection for local features. The entire network can be learned end-to-end by carefully balancing the gradient flow between two heads -- requiring only image-level labels. We also introduce an autoencoder-based dimensionality reduction technique for local features, which is integrated into the model, improving training efficiency and matching performance. Comprehensive experiments show that our model achieves state-of-the-art image retrieval on the Revisited Oxford and Paris datasets, and state-of-the-art single-model instance-level recognition on the Google Landmarks dataset v2. Code and models are available at https://github.com/tensorflow/models/tree/master/research/delf .