Learning grammars for architecture-specific facade parsing - Archive ouverte HAL Access content directly
Reports (Research Report) Year : 2014

Learning grammars for architecture-specific facade parsing

Abstract

Parsing facade images requires optimal handcrafted grammar for a given class of buildings. Such a handcrafted grammar is often designed manually by experts. In this paper, we present a novel framework to learn a compact grammar from a set of ground-truth images. To this end, parse trees of ground-truth annotated images are obtained running existing inference algorithms with a simple, very general grammar. From these parse trees, repeated subtrees are sought and merged together to share derivations and produce a grammar with fewer rules. Furthermore, unsupervised clustering is performed on these rules, so that, rules corresponding to the same complex pattern are grouped together leading to a rich compact grammar. Experimental validation and comparison with the state-of-the-art grammar-based methods on four diff erent datasets show that the learned grammar helps in much faster convergence while producing equal or more accurate parsing results compared to handcrafted grammars as well as grammars learned by other methods. Besides, we release a new dataset of facade images from Paris following the Art-deco style and demonstrate the general applicability and extreme potential of the proposed framework.
Fichier principal
Vignette du fichier
paper.pdf (21.29 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01069379 , version 1 (29-09-2014)
hal-01069379 , version 2 (02-10-2015)

Identifiers

  • HAL Id : hal-01069379 , version 1

Cite

Raghudeep Gadde, Renaud Marlet, Nikos Paragios. Learning grammars for architecture-specific facade parsing. [Research Report] RR-8600, INRIA. 2014. ⟨hal-01069379v1⟩
899 View
890 Download

Share

Gmail Facebook X LinkedIn More