Compact gesture recognition algorithm using machine learning
Gesture recognition is an important human-computer interaction tool that has been studied since the 1980s. From the very beginning with data gloves, gesture recognition has evolved to machine learning based gesture recognition, and the accuracy and application of gesture recognition has increased dr...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Coursework |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/155038 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Gesture recognition is an important human-computer interaction tool that has been studied since the 1980s. From the very beginning with data gloves, gesture recognition has evolved to machine learning based gesture recognition, and the accuracy and application of gesture recognition has increased dramatically. In this work,the classical LeNet-5 algorithm architecture is used. By tuning different parameters in Windows Caffe platform, 95% accuracy is obtained for
the recognition of gesture numbers from 0 to 5. The hardware is implemented in FPGA through Vivado design kit, and finally real time gesture display and result output is achieved. |
|---|