Semantically Tied Paired Cycle Consistency for Any-Shot Sketch-based Image Retrieval

TL;DR

This paper introduces SEM-PCYC, a novel generative adversarial network that enables zero-shot and few-shot sketch-based image retrieval by mapping sketches and images into a shared semantic space without requiring aligned pairs.

Contribution

The paper proposes a semantically aligned paired cycle-consistent GAN for any-shot SBIR, incorporating class-specific supervision and combining textual and hierarchical information for improved retrieval.

Findings

Significant performance boost over state-of-the-art on multiple datasets.

Effective handling of zero-shot and few-shot SBIR tasks.

Avoids the need for costly aligned sketch-image pairs.

Abstract

Low-shot sketch-based image retrieval is an emerging task in computer vision, allowing to retrieve natural images relevant to hand-drawn sketch queries that are rarely seen during the training phase. Related prior works either require aligned sketch-image pairs that are costly to obtain or inefficient memory fusion layer for mapping the visual information to a semantic space. In this paper, we address any-shot, i.e. zero-shot and few-shot, sketch-based image retrieval (SBIR) tasks, where we introduce the few-shot setting for SBIR. For solving these tasks, we propose a semantically aligned paired cycle-consistent generative adversarial network (SEM-PCYC) for any-shot SBIR, where each branch of the generative adversarial network maps the visual information from sketch and image to a common semantic space via adversarial training. Each of these branches maintains cycle consistency that only requires supervision at the category level, and avoids the need of aligned sketch-image pairs. A classification criteria on the generators' outputs ensures the visual to semantic space mapping to be class-specific. Furthermore, we propose to combine textual and hierarchical side information via an auto-encoder that selects discriminating side information within a same end-to-end model. Our results demonstrate a significant boost in any-shot SBIR performance over the state-of-the-art on the extended version of the challenging Sketchy, TU-Berlin and QuickDraw datasets.