Consiglio Nazionale delle Ricerche

Tipo di prodottoArticolo in rivista
TitoloVirtual to real adaptation of pedestrian detectors
Anno di pubblicazione2020
FormatoElettronico
Autore/iCiampi L.; Messina N.; Falchi F.; Gennaro C.; Amato G.
Affiliazioni autoriCNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy
Autori CNR e affiliazioni
  • NICOLA MESSINA
  • LUCA CIAMPI
  • GIUSEPPE AMATO
  • CLAUDIO GENNARO
  • FABRIZIO FALCHI
Lingua/e
  • inglese
AbstractPedestrian detection through Computer Vision is a building block for a multitude of applications. Recently, there has been an increasing interest in convolutional neural network-based architectures to execute such a task. One of these supervised networks' critical goals is to generalize the knowledge learned during the training phase to new scenarios with different characteristics. A suitably labeled dataset is essential to achieve this purpose. The main problem is that manually annotating a dataset usually requires a lot of human effort, and it is costly. To this end, we introduce ViPeD (Virtual Pedestrian Dataset), a new synthetically generated set of images collected with the highly photo-realistic graphical engine of the video game GTA V (Grand Theft Auto V), where annotations are automatically acquired. However, when training solely on the synthetic dataset, the model experiences a Synthetic2Real domain shift leading to a performance drop when applied to real-world images. To mitigate this gap, we propose two different domain adaptation techniques suitable for the pedestrian detection task, but possibly applicable to general object detection. Experiments show that the network trained with ViPeD can generalize over unseen real-world scenarios better than the detector trained over real-world data, exploiting the variety of our synthetic dataset. Furthermore, we demonstrate that with our domain adaptation techniques, we can reduce the Synthetic2Real domain shift, making the two domains closer and obtaining a performance improvement when testing the network over the real-world images.
Lingua abstractinglese
Altro abstract-
Lingua altro abstract-
Pagine da-
Pagine a-
Pagine totali14
RivistaSensors (Basel)
Attiva dal 2001
Editore: Molecular Diversity Preservation International (MDPI), - Basel
Lingua: inglese
ISSN: 1424-8220
Titolo chiave: Sensors (Basel)
Titolo proprio: Sensors. (Basel)
Titolo abbreviato: Sensors (Basel)
Numero volume della rivista20
Fascicolo della rivista18
DOI10.3390/s20185250
Verificato da refereeSì: Internazionale
Stato della pubblicazionePublished version
Indicizzazione (in banche dati controllate)
  • Scopus (Codice:2-s2.0-85090866382)
  • ISI Web of Science (WOS) (Codice:000586556700001)
Parole chiavePedestrian detection, Domain adaptation, Synthetic datasets, Convolutional neural networks, Deep learning
Link (URL, URI)https://www.mdpi.com/1424-8220/20/18/5250
Titolo parallelo-
Licenza-
Scadenza embargo-
Data di accettazione-
Note/Altre informazioni-
Strutture CNR
  • ISTI — Istituto di scienza e tecnologie dell'informazione "Alessandro Faedo"
Moduli/Attività/Sottoprogetti CNR
  • ICT.P08.010.002 : Digital Libraries
Progetti Europei-
Allegati
Virtual to Real Adaptation of Pedestrian Detectors
Descrizione: published versione OA
Tipo documento: application/pdf