Pseudo-LiDAR-based methods for monocular 3D object detection have received considerable attention in the community due to the performance gains exhibited on the KITTI3D benchmark, in particular on the commonly reported validation split. This generated a distorted impression about the superiority of Pseudo-LiDAR-based (PL-based) approaches over methods working with RGB images only. Our first contribution consists in rectifying this view by pointing out and showing experimentally that the validation results published by PL-based methods are substantially biased. The source of the bias resides in an overlap between the KITTI3D object detection validation set and the training/validation sets used to train depth predictors feeding PL-based methods. Surprisingly, the bias remains also after geographically removing the overlap. This leaves the test set as the only reliable set for comparison, where published PL-based methods do not excel. Our second contribution brings PL-based methods back up in the ranking with the design of a novel deep architecture which introduces a 3D confidence prediction module. We show that 3D confidence estimation techniques derived from RGB-only 3D detection approaches can be successfully integrated into our framework and, more importantly, that improved performance can be obtained with a newly designed 3D confidence measure, leading to state-of-the-art performance on the KITTI3D benchmark.

Are we Missing Confidence in Pseudo-LiDAR Methods for Monocular 3D Object Detection? / Simonelli, Andrea; Rota Bulo, Samuel; Porzi, Lorenzo; Kontschieder, Peter; Ricci, Elisa. - ELETTRONICO. - (2021), pp. 3205-3213. (Intervento presentato al convegno ICCV tenutosi a Montreal, QC, Canada nel 11-17 Ottobre, 2021) [10.1109/ICCV48922.2021.00321].

Are we Missing Confidence in Pseudo-LiDAR Methods for Monocular 3D Object Detection?

Simonelli, Andrea;Rota Bulo, Samuel;Ricci, Elisa
2021-01-01

Abstract

Pseudo-LiDAR-based methods for monocular 3D object detection have received considerable attention in the community due to the performance gains exhibited on the KITTI3D benchmark, in particular on the commonly reported validation split. This generated a distorted impression about the superiority of Pseudo-LiDAR-based (PL-based) approaches over methods working with RGB images only. Our first contribution consists in rectifying this view by pointing out and showing experimentally that the validation results published by PL-based methods are substantially biased. The source of the bias resides in an overlap between the KITTI3D object detection validation set and the training/validation sets used to train depth predictors feeding PL-based methods. Surprisingly, the bias remains also after geographically removing the overlap. This leaves the test set as the only reliable set for comparison, where published PL-based methods do not excel. Our second contribution brings PL-based methods back up in the ranking with the design of a novel deep architecture which introduces a 3D confidence prediction module. We show that 3D confidence estimation techniques derived from RGB-only 3D detection approaches can be successfully integrated into our framework and, more importantly, that improved performance can be obtained with a newly designed 3D confidence measure, leading to state-of-the-art performance on the KITTI3D benchmark.
2021
2021 IEEE/CVF International Conference on Computer Vision (ICCV)
Piscataway, NJ USA
IEEE/CVF International Conference on Computer Vision
978-1-6654-2812-5
Simonelli, Andrea; Rota Bulo, Samuel; Porzi, Lorenzo; Kontschieder, Peter; Ricci, Elisa
Are we Missing Confidence in Pseudo-LiDAR Methods for Monocular 3D Object Detection? / Simonelli, Andrea; Rota Bulo, Samuel; Porzi, Lorenzo; Kontschieder, Peter; Ricci, Elisa. - ELETTRONICO. - (2021), pp. 3205-3213. (Intervento presentato al convegno ICCV tenutosi a Montreal, QC, Canada nel 11-17 Ottobre, 2021) [10.1109/ICCV48922.2021.00321].
File in questo prodotto:
File Dimensione Formato  
Simonelli_Are_We_Missing_Confidence_in_Pseudo-LiDAR_Methods_for_Monocular_3D_ICCV_2021_paper.pdf

accesso aperto

Descrizione: Computer Vision Foundation version
Tipologia: Versione editoriale (Publisher’s layout)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 1.86 MB
Formato Adobe PDF
1.86 MB Adobe PDF Visualizza/Apri
Are_we_Missing_Confidence_in_Pseudo-LiDAR_Methods_for_Monocular_3D_Object_Detection.pdf

Solo gestori archivio

Tipologia: Versione editoriale (Publisher’s layout)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 2.33 MB
Formato Adobe PDF
2.33 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/11572/330534
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 7
  • ???jsp.display-item.citation.isi??? 3
social impact