Deep Factorised Inverse-Sketching

TL;DR

This paper introduces an unsupervised style transfer model that inverts free-hand sketches into geometric contours and separates salient details, improving sketch-photo matching in fine-grained image retrieval.

Contribution

It presents a novel unsupervised style transfer approach with cyclic embedding consistency and a factorised sketch representation for enhanced sketch-based image retrieval.

Findings

Outperforms state-of-the-art style transfer methods

Achieves better sketch-photo matching accuracy

Demonstrates qualitative and quantitative improvements

Abstract

Modelling human free-hand sketches has become topical recently, driven by practical applications such as fine-grained sketch based image retrieval (FG-SBIR). Sketches are clearly related to photo edge-maps, but a human free-hand sketch of a photo is not simply a clean rendering of that photo's edge map. Instead there is a fundamental process of abstraction and iconic rendering, where overall geometry is warped and salient details are selectively included. In this paper we study this sketching process and attempt to invert it. We model this inversion by translating iconic free-hand sketches to contours that resemble more geometrically realistic projections of object boundaries, and separately factorise out the salient added details. This factorised re-representation makes it easier to match a free-hand sketch to a photo instance of an object. Specifically, we propose a novel unsupervised image style transfer model based on enforcing a cyclic embedding consistency constraint. A deep FG-SBIR model is then formulated to accommodate complementary discriminative detail from each factorised sketch for better matching with the corresponding photo. Our method is evaluated both qualitatively and quantitatively to demonstrate its superiority over a number of state-of-the-art alternatives for style transfer and FG-SBIR.