Anup Nandy

Recognizing & interpreting Indian Sign Language gesture for Human Robot Interaction

This paper describes a novel approach towards recognizing of Indian Sign Language (ISL) gestures for Humanoid Robot Interaction (HRI). An extensive approach is being introduced for classification of ISL gesture which imparts an elegant way of interaction between humanoid robot HOAP-2 and human being. ISL gestures are being considered as a communicating agent for humanoid robot which is being used in this context explicitly. It involves different image processing techniques followed by a generic algorithm for feature extraction process. The classification technique deals with the Euclidean distance metric. The concrete HRI system has been established for initiation based learning mechanism. The Real time robotics simulation software, WEBOTS has been adopted to simulate the classified ISL gestures on HOAP-2 robot. The JAVA based software has been developed to deal with the entire HRI process.

Recognition of Isolated Indian Sign Language Gesture in Real Time

Indian Sign Language (ISL) consists of static as well as dynamic hand gestures for communication among deaf and dumb persons. Most of the ISL gestures are produced using both hands. A video database is created and utilized which contains several videos, for a large number of signs. Direction histogram is the feature used for classification due to its appeal for illumination and orientation invariance. Two different approaches utilized for recognition are Euclidean distance and K-nearest neighbor metrics.

A Framework for Synthesis of Human Gait Oscillation Using Intelligent Gait Oscillation Detector (IGOD)

The main objective of this paper illustrates an elementary concept about the designing, development and implementation of a bio-informatics diagnostic tool which understands and analyzes the human gait oscillation in order to provide an insight on human bi-pedal locomotion and its stability. A multi sensor device for detection of gait oscillations during human locomotion has been developed effectively. It has been named “IGOD”, an acronym of the “Intelligent Gait Oscillation Detector”. It ensures capturing of different person’s walking pattern in a very elegant way. This device would be used for creating a database of gait oscillations which could be extensively applied in several implications. The preliminary acquired data for eight major joints of a human body have been presented significantly. The electronic circuit has been attached to IGOD device in order to customize the proper calibration of every joint angle eventually.

Cloth invariant gait recognition using pooled segmented statistical features

Abstract A natural and normal gait can be used as a biometric cue in finding a solution to the human identification problem. An individual׳s appearance is likely to change with the variation in different clothes which further compounds the problem of gait identification. The clothing differences between gallery and probe datasets capture the possible changes in their silhouette׳s shape which increases the inability to discriminate between individuals. In this paper, an attempt has been made to provide a novel statistical shape analysis method based on Gait Energy Image (GEI) which is decomposed into three independent shape segmentations such as horizontal, vertical and grid resolution. The pooled segmented statistical features describe the shape of the GEI edge contour. The higher order moments about the shape centroid are likely to be invariant to small changes in silhouette shape. They implicitly describe the underlying distribution of the shape and can be used in conjunction with a set of other area based features to increase the efficacy of the classification results. The features reliability test has been performed with three classical statistical methods such as intra cloths variance (F-Statistics), inter subject distance (t-Statistics) and Intra-Class Correlation (ICC) on each set of segment of features. This analysis illustrates that combination of features holds less discrimination in comparison to grid based shape segmentation for different clothes. The similarity measurement comprises of different classification techniques (k-Nearest Neighbor, Naive Bayes׳, Decision Tree (C4.5) and Random Forest) to produce acceptable recognition results on OU-ISIR dataset. The degree of discriminability of these classifiers has been measured by statistical metrics such as F1-Score, Precision, Recall, and ROC curve.

Development of a Robust Microcontroller Based Intelligent Prosthetic Limb

Adaptive Modular Active Leg (AMAL), a robotic Intelligent Prosthetic Limb has been developed at the Indian Institute of Information Technology Allahabad. The aim of the project was to provide the comfort of an intelligent prosthetic knee joint for differently abeled person with one leg amputated above the knee. AMAL provides him with the necessary shock absorption and a suitable bending of the knee joint oscillation. The bending and the shock absorption are provided by artificial muscles. In our case, it is the MR (Magneto Rheological) damper which controls the knee movement of an amputee. The feedback signal is provided by the heel’s strike sensor. AMAL has been kept simple with minimal feedback sensors and controls so that the product is economically viable for the patients. In this paper we describe the mechanical design, the electronic control with its successful testing on differently abeled persons.

Modeling a Central Pattern Generator to Generate the Biped Locomotion of a Bipedal Robot Using Rayleigh Oscillators

This paper mainly deals with designing a biological controller for biped robot to generate biped locomotion inspired from human gait oscillation. The nonlinear dynamics of the biological controller is modeled by designing a Central Pattern Generator (CPG) which is the coupling of the Relaxation Oscillators. In this work the CPG consists of four Two-Way coupled Rayleigh Oscillators. The four major leg joints (e.g. two knee joints and two hip joints) are being considered for this modeling. The CPG parameters are optimized using Genetic Algorithm (GA) to match an actual human locomotion captured by the Intelligent Gait Oscillation Detector (IGOD) biometric device. The Limit Cycle behavior and the dynamic analysis on the biped robot have been successfully simulated on Spring Flamingo robot in YOBOTICS environment.

Gait Biometrics: An Approach to Speed Invariant Human Gait Analysis for Person Identification

A simple and a common human gait can provide an interesting behavioral biometric feature for robust human identification. The human gait data can be obtained without the subject’s knowledge through remote video imaging of people walking. In this paper we apply a computer vision-based technique to identify a person at various walking speeds, varying from 2 km/hr to 10 km/hr. We attempt to construct a speed invariance human gait classifier. Gait signatures are derived from the sequence of silhouette frames at different gait speeds. The OU-ISIR Treadmill Gait Databases has been used. We apply a dynamic edge orientation histogram on silhouette images at different speeds, as feature vector for classification. This orientation histogram offers the advantage of accumulating translation and orientation invariant gait signatures. This leads to a choice of the best features for gait classification. A statistical technique based on Naive Bayesian approach has been applied to classify the same person at different gait speeds. The classifier performance has been evaluated by estimating the maximum likelihood of occurrences of the subject.

A Sensor-Based Technique for Speed Invariant Human Gait Classification

Analyzing the human gait and obtaining the walking patterns can be an important biometric signature through which one could confirm an individual’s identity. In this paper, a nonvision-based approach using rotation sensor has been applied to acquire the oscillations from eight major joints of human body. These joints are, both the shoulders, elbows which constitute the upper body, and both hips and knees, which constitute the lower body. The gait patterns (from these eight oscillations) for male and female were obtained for different gait speeds varying from 3 to 5 km/h. The 3-km/h data was used as reference gait speed for training to classify the data at other gait speeds (4 and 5 km/h). This speed invariant human gait classification was done using a naive Bayesian classifier along with applying Euclidean distance method and K-nearest neighbor technique. We have achieved encouraging classification results with those techniques.

A study on gait entropy image analysis for clothing invariant human identification

A simple and common human gait may be viewed as a strong biometric cue to solve human identification problem through understanding the intrinsic patterns of gait biometrics. An individual’s gait pattern appears to be different in gallery and probe gait sequences due to wearing dissimilar clothing types. The gait dataset captures the possible changes found in silhouette shape image which provides the difficulty in distinguishing among individuals. In this paper, a robust feature selection technique has been addressed through Gait Entropy Image (GEnI) analysis. The GEnI has the capacity to accumulate most significant motion information. The width of GEnI, along the horizontal axis is taken as discriminative feature which produces a small intra-class variance. This information is studied as an evidence of feature invariance. The standard statistical tests such as pair-wise clothing correlation and intra-clothing variance are performed on gait dataset to evaluate the reliability of feature. Experimental results demonstrate the efficiency of proposed feature selection method using k-nearest neighbor (k-NN), minimum distance classifier (MDC), and support vector machine (SVM) algorithms. The performance analysis of recognition system has been evaluated on OU-ISIR Treadmill B gait database with different error metrics after performing N-fold cross validation method.

A new paradigm of human gait analysis with Kinect

Analysis of human gait helps to find an intrinsic gait signature through which ubiquitous human identification and medical disorder problems can be investigated in a broad spectrum. The gait biometric provides an unobtrusive feature by which video gait data can be captured at a larger distance without prior awareness of the subject. In this paper, a new technique has been addressed to study the human gait analysis with Kinect Xbox device. It ensures us to minimize the segmentation errors with automated background subtraction technique. The closely similar human skeleton model can be generated from background subtracted gait images, altered by covariate conditions, such as change in walking speed and variations in clothing type. The gait signatures are captured from joint angle trajectories of left hip, left knee, right hip and right knee of subjects skeleton model. The experimental verification on Kinect gait data has been compared with our in-house development of sensor based biometric suit, Intelligent Gait Oscillation Detector (IGOD). An endeavor has been taken to investigate whether this sensor based biometric suit can be altered with a Kinect device for the proliferation of robust gait identification system. The Fisher discriminant analysis has been applied on training gait signature to look into the discriminatory power of feature vector. The Naïve Bayesian classifier demonstrates an encouraging classification result with estimation of errors on limited dataset captured by Kinect sensor.