Zapisane w:
| 1. autor: | |
|---|---|
| Format: | Recurso digital |
| Język: | |
| Wydane: |
Zenodo
2025
|
| Dostęp online: | https://doi.org/10.5281/zenodo.15617010 |
| Etykiety: |
Dodaj etykietę
Nie ma etykietki, Dołącz pierwszą etykiete!
|
Spis treści:
- <p>In this paper, we try to recreate a simple gesture-recognition model attached to a camera that would then be used to interact with a computer such as clicking and scrolling on websites. In order to do this, we used mediapipe's basic online gesture recognition AI and colab's capture video code in order to capture images that would then be fed into the model to determine the gesture. Determining the gesture was hard for the AI, especially if the lighting or camera angle was off even a tiny bit. A way of interacting with websites has not been made due to the complications of implementing a way to do that. Overall, the model had a somewhat low accuracy that could intervene in interacting with computers that could cause it to do the wrong actions. In the future, better algorithms or data could be used to deal with these limitations.</p>