A Real-Time Hand Posture Recognition System Using Deep Neural Networks

Abstract

Hand posture recognition (HPR) is quite a challenging task, due to both the difficulty in detecting and tracking hands with normal cameras and the limitations of traditional manually selected features. In this article, we propose a two-stage HPR system for Sign Language Recognition using a Kinect sensor. In the first stage, we propose an effective algorithm to implement hand detection and tracking. The algorithm incorporates both color and depth information, without specific requirements on uniform-colored or stable background. It can handle the situations in which hands are very close to other parts of the body or hands are not the nearest objects to the camera and allows for occlusion of hands caused by faces or other hands. In the second stage, we apply deep neural networks (DNNs) to automatically learn features from hand posture images that are insensitive to movement, scaling, and rotation. Experiments verify that the proposed system works quickly and accurately and achieves a recognition accuracy as high as 98.12%. . 2015. A real-time hand posture recognition system using deep neural networks.

Sign Language Recognition using 3D convolutional neural networks

“…The framework is shown in Fig.5. We follow the idea proposed by [16] to crop hand-shape from background. [16] also used Kinect as input device and proposed an effective algorithm to implement hand segmentation and tracking.…”

“…We follow the idea proposed by [16] to crop hand-shape from background. [16] also used Kinect as input device and proposed an effective algorithm to implement hand segmentation and tracking. The algorithm incorporates both color and depth information, without specific requirements on uniform-colored or stable background.…”

Estimation of Hand Skeletal Postures by Using Deep Convolutional Neural Networks

“…Another way is to use optical sensors, because they provide superior tracking and detecting capabilities due to their high resolution [8]. In [9], both features of color and depth with a deep neural network are used. There are many other algorithms for hand segmentation in which some of them use the skin color feature, such as region growing algorithm [9][10][11].…”

“…In [9], both features of color and depth with a deep neural network are used. There are many other algorithms for hand segmentation in which some of them use the skin color feature, such as region growing algorithm [9][10][11]. Also, many different methods and algorithms perform hand segmentation by removing the background [12].…”

Reconhecimento de gestos em vídeos utilizando modelos ocultos de Markov e redes neurais convolucionais aplicado a LIBRAS

“…Outra técnica mais modernaé o uso de redes neurais convolucionais (CNNs) para reconhecer as posturas das mãos, comoé feito em [2]. As CNNs se provaraḿ otimos reconhecedores de padrões [3] [4] gerando resultados superiores a outras técnicas mais clássicas como SVM ou ANNs. Apesar das CNNs serem normalmente utilizadas para o reconhecimento de imagens estáticas, como a forma da mão, em [5] elas são utilizadas para fazer o reconhecimento de gestos dinâmicos isolados.…”