Yüksek Lisans Tez Sunumu: Learning Visual Similarity for Image Retrieval with Global Descriptors and Capsule Networks

Duygu Durmuş

Özet:

Finding matching images across large and unstructured datasets plays an important
in many computer vision applications. With the emergence of deep learningbased
solutions, various visual tasks such as image retrieval have been successfully
addressed. Learning visual similarity is crucial for image matching and retrieval
tasks. An alternative deep learning architecture, named capsule networks, enables
learning richer information that describes the object without losing the essential
spatial relationship between the object and its parts. Also, global descriptors are
widely used for representing images.

The proposed architecture combines the power of global descriptors and revised
capsule networks to enhance image retrieval performance. It benefits from multiple
views of object images and highlights the spatial relationship between objects
and their parts. Spatial Grouping Enhance strategy, which enhances sub-features
parallelly, and self-attention layers, which explore global dependencies within internal
representations of images, are utilized to empower the image representations.
The approach captures resemblances between similar images and differences
between the non-similar images using both triplet loss and cost-sensitive
regularized cross-entropy loss instead of learning classification for individual images.
Based on the experiments, the results are superior to the state-of-the-art
approaches for Stanford Online Products.