Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{205920,
        author = {Ketan Kumar Kaiwart and Pratik Vijay Manjarekar and Prasad Sidrameshwar Khanapure},
        title = {Real time cursor control using hand gestures with Vision Transformer},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {9116-9124},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205920},
        abstract = {This paper present real-time cursor control using hand gestures with vision transformer for contactless human computer interaction (HCI). The proposed system uses webcam for capturing live hand movement, while Media Pipe can efficient hand detect & region-of-interest (ROI) extraction. The extracted hand images are pre-processed and classified into gesture classes using a pretrained Vision Transformer model fine-tuned on a custom gesture dataset. The cursor actions such as movement, click, drag and scrolling are mapped using recognized gestures through PyAutoGUI. The system follows pipeline consist hand detection, preprocess the input data, gesture classification and cursor actions mapping. To improve performance in real time environment use smoothing and gesture debouncing techniques. The model is train using two phase transfer learning strategy, in that classification head is initially train with frozen backbone followed by full fine tuning of the model. Experimental evaluation Conduct on custom dataset content images across 7,248 for 7 gesture classes. The system achieves 99.91% classification accuracy on the static dataset and 93.52% accuracy during real-time, the accuracy in real time conditions get vary. The results demonstrate that Vision Transformers effectively capture global spatial relationships between hand features, providing a robust and efficient alternative to traditional input devices for contactless computer applications.},
        keywords = {Virtual Mouse, Gesture Recognition, Human-Computer Interaction (HCI), Computer Vision, CNN, Vision Transformer (ViT).},
        month = {June},
        }