Towards Object Re-identification from Point Clouds for 3D MOT
Abstract
This thesis studies the problem of object re-identification (ReID) in a 3D multi-object tracking (MOT) context, by learning to match pairs of objects from cropped (e.g., using their predicted 3D bounding boxes) point cloud observations. We are not concerned with state-of-the-art performance for 3D MOT, however. Instead, we seek to answer the following question: In a realistic tracking by-detection context, how does object ReID from point clouds perform relative to ReID from images? To enable such a study, we propose a lightweight matching head that can be concatenated to any set or sequence processing backbone (e.g., PointNet or ViT), creating a family of comparable object ReID networks for both modalities. Run in Siamese style, our proposed point cloud ReID networks can make thousands of pairwise comparisons in real-time (10 Hz). Our findings demonstrate that their performance increases with higher sensor resolution and approaches that of image ReID when observations are sufficiently dense. Additionally, we investigate our network's ability to enhance 3D multi-object tracking, showing that our point cloud ReID networks can successfully re-identify objects that led a strong motion-based tracker into error. To our knowledge, we are the first to study real-time object re-identification from point clouds in a 3D multi-object tracking context.
Collections
Cite this version of the work
Benjamin Thérien
(2023).
Towards Object Re-identification from Point Clouds for 3D MOT. UWSpace.
http://hdl.handle.net/10012/19305
Other formats
Related items
Showing items related by title, author, creator and subject.
-
Decentralized Ambient System Identification of Structures
Sadhu, Ayan (University of Waterloo, 2013-05-22)Many of the existing ambient modal identification methods based on vibration data process information centrally to calculate the modal properties. Such methods demand relatively large memory and processing capabilities to ... -
How We Became Legion: Burke's Identification and Anonymous
Ramos Antunes da Silva, Debora Cristina (University of Waterloo, 2013-08-19)This thesis presents a study of how identification, according to Kenneth Burke's theory, can be observed in the media-related practices promoted by the cyber-activist collective Anonymous. Identification is the capacity ... -
The Development of System Identification Approaches for Complex Haptic Devices and Modelling Virtual Effects Using Fuzzy Logic
Tam, Sze-Man Samantha (University of Waterloo, 2005)Haptic applications often employ devices with many degrees of freedom in order to allow the user to have natural movement during human-machine interaction. From the development point of view, the complexity in ...