Deciding of HMM parameters based on number of critical points for gesture recognition from motion capture data (original) (raw)

Comparison study of Hidden Markov Model gesture recognition using fixed state and variable state

2013 IEEE International Conference on Signal and Image Processing Applications, 2013

This paper presents a method of gesture recognition using Hidden Markov Model (HMM). Gesture itself is based on the movement of each right hand (RH) and left hand (LH), which represents the word intended by the signer. The feature vector selected, gesture path, hand distance and hand orientations are obtained from RH and LH then trained using HMM to produce the respective gesture class. While training, in handling HMM state, we introduce fixed state and variable state, where in fixed state, the numbers of state is generally fixed for all gestures and while the number of state in variable state is determined by the movement of the gesture. It was found that fixed state gave the highest rate of recognition achieving 83.1 %.

An HMM-based threshold model approach for gesture recognition

Pattern Analysis and Machine Intelligence, …, 1999

AbstractÐThe task of automatic gesture recognition is highly challenging due to the presence of unpredictable and ambiguous nongesture hand motions. In this paper, a new method is developed using the Hidden Markov Model based technique. To handle nongesture patterns, we ...

Model Structure Selection & Training Algorithms for a HMM Gesture Recognition System

Hidden Markov models using the Fully-Connected, Left-Right and Left-Right Banded model structures are applied to the problem of alphabetical letter gesture recognition. We examine the effect of training techniques, in particular the Baum-Welch and Viterbi Path Counting techniques, on each of the model structures. We show that recognition rates improve when moving from a Fully-Connected model to a Left-Right model and a Left-Right Banded 'staircase' model with peak recognition rates of 84.8%, 92.31% and 97.31% respectively. The Left-Right Banded model in conjunction with the Viterbi Path Counting present the best performance. Direct calculation of model parameters from analysis of the physical system was also tested, yielding a peak recognition rate of 92%, but the simplicity and efficiency of this approach is of interest.

Model structure selection & training algorithms for an HMM gesture recognition system

Frontiers in Handwriting …, 2004

Model Structure Selection and Training Algorithms for an HMM Gesture Recognition System

Ninth International Workshop on Frontiers in Handwriting Recognition, 2004

Hidden Markov model for human to computer interaction: a study on human hand gesture recognition

Artificial Intelligence Review, 2011

Human hand recognition plays an important role in a wide range of applications ranging from sign language translators, gesture recognition, augmented reality, surveillance and medical image processing to various Human Computer Interaction (HCI) domains. Human hand is a complex articulated object consisting of many connected parts and joints. Therefore, for applications that involve HCI one can find many challenges to establish a system with high detection and recognition accuracy for hand posture and/or gesture. Hand posture is defined as a static hand configuration without any movement involved. Meanwhile, hand gesture is a sequence of hand postures connected by continuous motions. During the past decades, many approaches have been presented for hand posture and/or gesture recognition. In this paper, we provide a survey on approaches which are based on Hidden Markov Models (HMM) for hand posture and gesture recognition for HCI applications.

Adaptation Procedure for HMM-Based Sensor-Dependent Gesture Recognition

Proceedings of the 8th ACM SIGGRAPH Conference on Motion in Games - SA '15, 2015

In this paper, we address the problem of sensor-dependent gesture recognition thanks to adaptation procedure. Capturing human movements by a motion capture (MoCap) system provides very accurate data. Unfortunately, such systems are very expensive, unlike recent depth sensors, like Microsoft Kinect, which are much cheaper, but provide lower data quality. Hidden Markov Models (HMMs) are widely used in gesture recognition to learn the dynamics of each gesture class. However, models trained on one type of data can only be used on data of the same type. For this reason, we propose to adapt HMMs trained on Mocap data to a small set of Kinect data using Maximum Likelihood Linear Regression (MLLR) to recognize gestures captured by a Kinect. Results show that using this method, we can achieve a recognition average accuracy of 84.48% using a small set of adaptation data while, using the same set to create new models, we obtain only 72.41% of accuracy.

Dynamic hand gesture recognition using hidden Markov models

Hand gesture recognition is one of the leading applications of human computer interaction. With diversity of applications of hand gesture recognition, sign language interpretation is the most demanding application. In this paper, dynamic hand gesture recognition for few subset of Indian sign language recognition was considered. The use of depth camera such as Kinect sensor gave skeleton information of signer body. After detailed study of dynamic ISL vocabulary with reference to skeleton joint information, angle has identified as a feature with reference to two moving hand. Here, real time video has been captured and gesture was recognized using Hidden Markov Model (HMM). Ten state HMM model was designed and normalized angle feature of dynamic sign was being observed. Maximum likelihood probability symbol was considered as a recognized gesture. Algorithm has been tested on ISL 20 dynamic signs of total 800 training set of four persons and achieved 89.25% average accuracy.

An HMM-Based Gesture Recognition Method Trained on Few Samples

2014 IEEE 26th International Conference on Tools with Artificial Intelligence, 2014

This paper addresses the problem of recognizing gestures which are captured using the Kinect sensor in a educational game devoted to the deaf community. Different strategies are evaluated to deal with the problem of having few samples for training. We have experimented a Leave One Out Training and Testing (LOOT) strategy and an HMM-based ensemble of classifiers. A dataset containing 181 videos of gestures related to nine signs commonly used in educational games is introduced, which is available for research purposes. The experimental results have shown that the proposed ensemble-based method is a promising strategy to deal with problems where few training samples are available.

Deciding of HMM parameters based on number of critical points for gesture recognition from motion capture data (original) (raw)

Related papers