CNN Based Posture-Free Hand Detection

TL;DR

This paper introduces a shallow CNN for hand detection that is fast, pose-invariant, and performs comparably to deeper models, making it suitable for real-time applications.

Contribution

A novel shallow CNN architecture for hand detection that balances speed and accuracy, reducing overfitting and computational complexity.

Findings

Achieves 93.9% accuracy on hand detection datasets.

Faster processing speed than existing CNN-based methods.

Insensitive to hand translation and pose variations.

Abstract

Although many studies suggest high performance hand detection methods, those methods are likely to be overfitting. Fortunately, the Convolution Neural Network (CNN) based approach provides a better way that is less sensitive to translation and hand poses. However the CNN approach is complex and can increase computational time, which at the end reduce its effectiveness on a system where the speed is essential.In this study we propose a shallow CNN network which is fast, and insensitive to translation and hand poses. It is tested on two different domains of hand datasets, and performs in relatively comparable performance and faster than the other state-of-the-art hand CNN-based hand detection method. Our evaluation shows that the proposed shallow CNN network performs at 93.9% accuracy and reaches much faster speed than its competitors.