Indoor UAV crowd investigation part 2 via computer vision applications and federated learning methods
In this report, the author examines different machine learning methods that aids in crowd counting in the novel context of a fixed location within NTU. The author aims to create an end-to-end solution by creating a self-made dataset and then testing it against contemporary ML models. As privacy is a...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
Nanyang Technological University
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/151006 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | In this report, the author examines different machine learning methods that aids in crowd counting in the novel context of a fixed location within NTU. The author aims to create an end-to-end solution by creating a self-made dataset and then testing it against contemporary ML models. As privacy is also a top concern that comes to mind for consumers, Federated Learning comes into play within this project. The author will conduct a quick treatment of which Federated algorithm should be used over the novel ones proposed by the scientific community. Lastly, the author attempts to convert the chosen crowd counting model into a mobile lite and federated model for the unique application within NTU. |
|---|