Multi-modality fusion in multiple object tracking

Tracking multiple individuals in unconstrained videos, especially in challenging scenarios like crowded environments or situations with visually similar people, presents substantial difficulties. Existing tracking methods heavily depend on sophisticated detectors and task-specific data association t...

Full description

Bibliographic Details
Main Author: Liao, Ruosen
Other Authors: Lin Zhiping
Format: Thesis-Master by Coursework
Language:English
Published: Nanyang Technological University 2024
Subjects:
Online Access:https://hdl.handle.net/10356/173003