Litcius/Paper detail

Cursor Movement Based on Object Detection Using Vision Transformers

Elizabeth Rani G, M Sakthimohan, Shiv Surya, S. Kalaiselvi, Tressa Bernice, V Gunasekran

202318 citationsDOI

Abstract

The research paper titled “Cursor Movement on Object Detection Using Vision Transformers” focuses on using Vision Transformers (ViT) to detect objects, specifically cursor movement. Modern deep learning architecture called ViT can spot things in still and moving pictures. The proposed method utilizes the transformer architecture to capture the contextual information of an image, which can be used to accurately detect and locate objects in the scene. Additionally, it uses a novel cursor movement algorithm that can accurately move the cursor toward the target object in real time. On common object identification datasets, the effectiveness of our suggested strategy is assessed and contrasted with numerous cutting-edge approaches. Experimental results demonstrate that our methodology outperforms state-of-the-art methods by a substantial margin.

Topics & Concepts

Computer visionComputer scienceCursor (databases)Artificial intelligenceObject detectionComputer graphics (images)Movement (music)Pattern recognition (psychology)PhilosophyAestheticsAdvanced Neural Network ApplicationsHuman Pose and Action RecognitionHand Gesture Recognition Systems