Type of Publication

Conference Papers

Date:

8 /

2020

Status

Preprint

DOI:

10.48550/arXiv.2005.14565

Probabilistic Object Classification using CNN ML-MAP layers

Featured in:

Workshop on Perception for Autonomous Driving (PAD), European Conference on Computer Vision (ECCV) 2020

Authors:

Gledson Melotti, Cristiano Premebida, Jordan Bird, Diego Faria and Nuno Gonçalves

Abstract

Deep networks are currently the state-of-the-art for sensory perception in autonomous driving and robotics. However, deep models often generate overconfident predictions precluding proper probabilistic interpretation which we argue is due to the nature of the SoftMax layer. To reduce the overconfidence without compromising the classification performance, we introduce a CNN probabilistic approach based on dis- tributions calculated in the network’s Logit layer. The approach enables Bayesian inference by means of ML and MAP layers. Experiments with calibrated and the proposed prediction layers are carried out on object classification using data from the KITTI database. Results are reported for camera (RGB) and LiDAR (range-view) modalities, where the new approach shows promising performance compared to SoftMax.

Citation
Gledson Melotti, Cristiano Premebida, Jordan Bird, Diego Faria and Nuno Gonçalves (2020). Probabilistic object classification using CNN ML-MAP layers. arXiv preprint arXiv:2005.14565.

Related Content

Researcher Coordinator, VIS TEAM Leader
PhD Student
No tagged content to show
No tagged content to show
No tagged content to show

RECENT PUBLICATIONS

VOIDFace: Towards an Effective Face Training Data Storage and Protection with Right to be Forgotten Property

Authors: Ajnas Muhammed; Iurii Medvedev; Nuno Gonçalves
Featured in: IEEE International Joint Conference on Biometrics (IJCB 2025)

Análise de marcadores biométricos e sinais vitais através de visão artificial: Ritmo Cardíaco

Authors: Rodrigo Rosado Gonçalves
Featured in: Bsc Thesis

How to Evaluate a Biometric Facial Image

Authors: Diana Gonçalves D’Amil
Featured in: Bsc Thesis

suggested news

Paper accepted to IJCB 2025
Prof. Nuno and VIS Team successfully organizes IbPRIA...
Four papers presented @ IbPRIA 2025

RECENT PROJECTS

FACING2 – Face Image Understanding
VISUAL-ID – Unique Visual Identities in Graphics, Images and Faces
UniqueMark

Institute of Systems and Robotics Department of Electrical and Computers Engineering University of Coimbra