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...
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis-Master by Coursework |
Language: | English |
Published: |
Nanyang Technological University
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/173003 |