Sketching out the details: Sketch-based image retrieval using convolutional neural networks with multi-stage regression

Author(s)

Bui, Tu ^[1] ; Ribeiro, Leonardo ^[2] ; Ponti, Moacir ^[2] ; Collomosse, John ^[1]

Affiliation

^[1] Univ Surrey, CVSSP, Guildford GU2 7XH, Surrey - England

^[2] Univ Sao Paulo, Inst Math & Comp Sci ICMC, BR-13566590 Sao Carlos, SP - Brazil

• Document type: Journal article
• Source: COMPUTERS & GRAPHICS-UK; v. 71, p. 77-87, APR 2018.
• Web of Science Citations: 3

Abstract

We propose and evaluate several deep network architectures for measuring the similarity between sketches and photographs, within the context of the sketch based image retrieval (SBIR) task. We study the ability of our networks to generalize across diverse object categories from limited training data, and explore in detail strategies for weight sharing, pre-processing, data augmentation and dimensionality reduction. In addition to a detailed comparative study of network configurations, we contribute by describing a hybrid multi-stage training network that exploits both contrastive and triplet networks to exceed state of the art performance on several SBIR benchmarks by a significant margin. (C) 2017 Elsevier Ltd. All rights reserved. (AU)

• FAPESP's process: 16/16111-4 - Feature learning applied to sketch-based image retrieval and low-altitude remote sensing
• Grantee: Moacir Antonelli Ponti
• Support type: Regular Research Grants
• FAPESP's process: 13/07375-0 - CeMEAI - Center for Mathematical Sciences Applied to Industry
• Grantee: José Alberto Cuminato
• Support type: Research Grants - Research, Innovation and Dissemination Centers - RIDC
• FAPESP's process: 17/10068-2 - Dimensionality reduction methods for representations generated by triplet convolutional networks
• Grantee: Leonardo Sampaio Ferraz Ribeiro
• Support type: Scholarships in Brazil - Scientific Initiation