Opening up open world tracking

Y Liu, IE Zulfikar, J Luiten, A Dave… - Proceedings of the …, 2022 - openaccess.thecvf.com
Tracking and detecting any object, including ones never-seen-before during model training,
is a crucial but elusive capability of autonomous systems. An autonomous agent that is blind …

Better call sal: Towards learning to segment anything in lidar

A Ošep, T Meinhardt, F Ferroni, N Peri… - … on Computer Vision, 2024 - Springer
We propose the SAL (S egment A nything in L idar) method consisting of a text-promptable
zero-shot model for segmenting and classifying any object in Lidar, and a pseudo-labeling …

Lidar Panoptic Segmentation in an Open World

AS Chakravarthy, MR Ganesina, P Hu… - International Journal of …, 2024 - Springer
Abstract Addressing Lidar Panoptic Segmentation (LPS) is crucial for safe deployment of
autnomous vehicles. LPS aims to recognize and segment lidar points wrt a pre-defined …

SSAT: Self-supervised associating network for multiobject tracking

TY Chung, MA Cho, H Lee, S Lee - IEEE Transactions on …, 2022 - ieeexplore.ieee.org
Multi-object tracking (MOT), which is crucial for computer vision and video processing, has
immense potential for improvement. Traditional tracking-by-detection approaches include …