3D-guided multiscale sliding window for pedestrian detection

Alejandro González, Gabriel Villalonga, German Ros, David Vázquez, Antonio M. López

Producció científica: Capítol de llibreContribució a congrés/conferènciaAvaluat per experts

5 Cites (Scopus)

Resum

The most relevant modules of a pedestrian detector are the candidate generation and the candidate classification. The former aims at presenting image windows to the latter so that they are classified as containing a pedestrian or not. Much attention has being paid to the classification module, while candidate generation has mainly relied on (multiscale) sliding window pyramid. However, candidate generation is critical for achieving real-time. In this paper we assume a context of autonomous driving based on stereo vision. Accordingly, we evaluate the effect of taking into account the 3D information (derived from the stereo) in order to prune the hundred of thousands windows per image generated by classical pyramidal sliding window. For our study we use a multimodal (RGB, disparity) and multi-descriptor (HOG, LBP, HOG+LBP) holistic ensemble based on linear SVM. Evaluation on data from the challenging KITTI benchmark suite shows the effectiveness of using 3D information to dramatically reduce the number of candidate windows, even improving the overall pedestrian detection accuracy.

Idioma originalAnglès
Títol de la publicacióPattern Recognition and Image Analysis - 7th Iberian Conference, IbPRIA 2015, Proceedings
EditorsJaime S. Cardoso, Roberto Paredes, Xosé M. Pardo
EditorSpringer Verlag
Pàgines560-568
Nombre de pàgines9
ISBN (electrònic)9783319193892
DOIs
Estat de la publicacióPublicada - 2015
Publicat externament
Esdeveniment7th Iberian Conference on Pattern Recognition and Image Analysis, IbPRIA 2015 - Santiago de Compostela, Spain
Durada: 17 de juny 201519 de juny 2015

Sèrie de publicacions

NomLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volum9117
ISSN (imprès)0302-9743
ISSN (electrònic)1611-3349

Conferència

Conferència7th Iberian Conference on Pattern Recognition and Image Analysis, IbPRIA 2015
País/TerritoriSpain
CiutatSantiago de Compostela
Període17/06/1519/06/15

Fingerprint

Navegar pels temes de recerca de '3D-guided multiscale sliding window for pedestrian detection'. Junts formen un fingerprint únic.

Com citar-ho