Deep Learning requires huge amount of data with related labels, that are necessary for proper training. Thanks to modern videogames, which aim at photorealism, it is possible to easily obtain syn- thetic dataset by extracting information directly from the game engine. The intent is to use data extracted from a videogame to obtain a repre- sentation of various scenarios and train a deep neural network to infer the information required for a specific task. In this work we focus on com- puter vision aids for automotive applications and we target to estimate the distance and speed of the surrounding vehicles by using a single dash- board camera. We propose two network models for distance and speed estimation, respectively. We show that training them by using synthetic images generated by a game engine is a viable solution that turns out to be very effective in real settings.

Estimation of speed and distance of surrounding vehicles from a single camera

Zaffaroni M.;Grangetto M.;
2019-01-01

Abstract

Deep Learning requires huge amount of data with related labels, that are necessary for proper training. Thanks to modern videogames, which aim at photorealism, it is possible to easily obtain syn- thetic dataset by extracting information directly from the game engine. The intent is to use data extracted from a videogame to obtain a repre- sentation of various scenarios and train a deep neural network to infer the information required for a specific task. In this work we focus on com- puter vision aids for automotive applications and we target to estimate the distance and speed of the surrounding vehicles by using a single dash- board camera. We propose two network models for distance and speed estimation, respectively. We show that training them by using synthetic images generated by a game engine is a viable solution that turns out to be very effective in real settings.
2019
20th International Conference on Image Analysis and Processing, ICIAP 2019
ita
2019
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Springer Verlag
11751
388
398
978-3-030-30641-0
978-3-030-30642-7
https://www.springer.com/series/558
Automotive; Computer vision; Deep Learning; Synthetic dataset
Zaffaroni M.; Grangetto M.; Farasin A.
File in questo prodotto:
File Dimensione Formato  
Camera Ready paper.pdf

Accesso aperto

Tipo di file: POSTPRINT (VERSIONE FINALE DELL’AUTORE)
Dimensione 1.51 MB
Formato Adobe PDF
1.51 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2318/1719738
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact