Identification of humans using gait

By Amit A. Kale, Aravind Sundaresan, Ambasamudram Narayanan Rajagopalan, Naresh P. Cuntoor, Amit K. Roy-Chowdhury, Volker Krüger, Ramalingam Chellappa. IEEE Transactions on Image Processing, 2004

647 citations recorded by the source corpus.

We propose a view-based approach to recognize humans from their gait. Two different image features have been considered: the width of the outer contour of the binarized silhouette of the walking person and the entire binary silhouette itself. To obtain the observation vector from the image features, we employ two different methods. In the first method, referred to as the indirect approach, the high-dimensional image feature is transformed to a lower dimensional space by generating what we call the frame to exemplar (FED) distance. The FED vector captures both structural and dynamic traits of each individual. For compact and effective gait representation and recognition, the gait information in the FED vector sequences is captured in a hidden Markov model (HMM). In the second method, referred to as the direct approach, we work with the feature vector directly (as opposed to computing the FED) and train an HMM. We estimate the HMM parameters (specifically the observation probability B) based on the distance between the exemplars and the image features. In this way, we avoid learning high-dimensional probability density functions. The statistical nature of the HMM lends overall robustness to representation and recognition. The performance of the methods is illustrated using several databases.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/5480518ea7a41f8feed52e963b6eb4efcc1142ac

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Behavioural biometrics: a survey and classification (2008, background): “Gait is one of the best researched muscle control-based biometrics (Benabdelkader et al., 2002; Kale et al., 2004; Nixon and Carter, 2004), it is a complex spatio-temporal motor-control behaviour which allows biometric recognition of individuals at a distance usually from captured video.”
Ieee Transactions on Pattern Analysis and Machine Intelligence 1 Matching Shape Sequences in Video with Applications in Human Movement Analysis. Ieee Transactions on Pattern Analysis and Machine Intelligence 2 (neutral): “Several researchers [38][39] have used Hidden Markov Models (HMM) for the task of gait based identification.”
Gait recognition: a challenging signal processing technology for biometric identification (2005, methodology): “For the above reasons, stochastic approaches such as HMMs [47] can also be used for gait recognition [32], [48].”
Human Action Segmentation and Recognition Using Discriminative Semi-Markov Models (2011, background): “Traditionally generative statistical approaches, especially the Markov models (Brand et al. 1997; Kale et al. 2004; Lv and Nevatia 2006; Yamato et al. 1992) have been in wide use to model and analyze human actions, e.”
General Tensor Discriminant Analysis and Gabor Features for Gait Recognition (2007, methodology): “Currently, there are several successful representation models such as appearance-based models [6], [16], [19], [22], [29], [30], [34], [37], [40], stochastic statistical models [ 19 ], articulated biomechanics models [36], [3], [1], [7] in which a set of parameters describes the gait, and other para”
Human Identification Using Temporal Information Preserving Gait Template (2012, methodology): “…“Rank1” and “Rank5” performances of several recent approaches including baseline algorithm (based on silhouette shape matching) [8], GEI [1], HMM [29], IMED+LDA [11], 2DLDA [11], DATER [10], MTP [12], Tensor Locality Preserving Projections (TLPP) [30], and Dynamics-Normalized Gait Recognition…”
Improved gait recognition by gait dynamics normalization (2006, methodology): “We consider recognition rates reported on publicly available experimental protocols and data sets (> 25 persons) such as the CMU-Mobo data set [5] (indoor, 25 subjects), UMD data set [ 6 ] (outdoor, 55 subjects), SOTON Large data set [7] (indoor and outdoor, 115 subjects), HumanID Gait Challenge dat”
Gait recognition using Gait Entropy Image (2009, methodology): “It can be seen from Table 4 that our approach outperforms the three model-free approaches, namely baseline [11], MSCT+SST [7], and GEI [3], and comparable to a HMM based approach [5].”
Machine Recognition of Human Activities: A Survey (2008, neutral): “[72], Liu and Sarkar [73]).”
MPCA: Multilinear Principal Component Analysis of Tensor Objects (2008, methodology): “Other heuristic features, such as widths of the outer contour [6] and angular transforms [4], have been shown to be poorer than silhouettes so they are not compared here.”
Gait Components and Their Application to Gender Recognition (2008, background): “Many other efforts have also been spent on gait information processing [2], [4], [6], [9], [10], [12], [19], [22], [38], and since then, gait silhouette information has been widely utilized, e.”