Variational Voxel Pseudo Image Tracking

Illia Oleksiienko, Paraskevi Nousi, Nikolaos Passalis, Anastasios Tefas, Alexandros Iosifidis

Research output: Contribution to book/anthology/report/proceedingArticle in proceedingsResearchpeer-review


Uncertainty estimation is an important task for critical problems, such as robotics and autonomous driving, because it allows creating statistically better perception models and signaling the model's certainty in its predictions to the decision method or a human supervisor. In this paper, we propose a Variational Neural Network-based version of a Voxel Pseudo Image Tracking (VPIT) method for 3D Single Object Tracking. The Variational Feature Generation Network of the proposed Variational VPIT computes features for target and search regions and the corresponding uncertainties, which are later combined using an uncertainty-aware cross-correlation module in one of two ways: by computing similarity between the corresponding uncertainties and adding it to the regular cross-correlation values, or by penalizing the uncertain feature channels to increase influence of the certain features. In experiments, we show that both methods improve tracking performance, while penalization of uncertain features provides the best uncertainty quality.
Original languageEnglish
Title of host publication2023 IEEE Symposium Series on Computational Intelligence (SSCI)
Number of pages6
Publication date2023
ISBN (Electronic)9781665430654
Publication statusPublished - 2023
SeriesProceedings (IEEE Symposium Series on Computational Intelligence)


  • 3D Single Object Tracking
  • Bayesian Neural Networks
  • Machine Learning for Embedded Devices
  • Point Cloud
  • Uncertainty Estimation
  • Variational Neural Networks


Dive into the research topics of 'Variational Voxel Pseudo Image Tracking'. Together they form a unique fingerprint.

Cite this