Generative networks and feature learning for cross domain visual search

• Grant number: 17/22366-8
• Support type: Scholarships in Brazil - Doctorate (Direct)
• Effective date (Start): March 01, 2018
• Effective date (End): February 28, 2023
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal researcher: Moacir Antonelli Ponti
• Grantee: Leonardo Sampaio Ferraz Ribeiro
• Home Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated scholarship(s): 19/02808-1 - Finding correspondences between cross-domain representations using unsupervised feature matching, BE.EP.DD

Abstract

Feature learning methods have reached state of the art on many areas of study. Despite excellent results achieved on benchmark datasets, there is still little understanding of its inner workings, and applications to be explored, particularly when considering architectures that go beyond traditional convolutional neural networks. In this project, we propose the use of feature learning for applications that encompass domain mapping, specifically applications that regard image retrieval with queries from different domains and applications with a limited amount of labelled data. These challenges can be tackled using deep learning, with the development of novel architectures based on multi-stream networks, convolutional autoencoders and generative models. The expected results include the development of models that can successfully map between different domains and that are also capable of generalizing to unseen data. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DOS SANTOS, FERNANDO P.; RIBEIRO, LEONARDO S. F.; PONTI, MOACIR A. Generalization of feature embeddings transferred from different video anomaly detection domains. JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION, v. 60, p. 407-416, APR 2019. Web of Science Citations: 3.