An in-depth analysis of the state of the art in gait-based human identification is provided in this book. Gait, that is, “the manner of walking,” can be construed as a differentiating factor among humans, and has been considered as a possible biometric measure able to identify and authenticate individuals. Given this perspective, it is natural to find this volume as the fourth installment of a larger international series on biometrics from this publisher. It is also commendable to the see the topic presented in a self-standing volume, and not included as a section in a more general biometrics handbook. The authors, coming from three different continents, exemplify the current international cooperation needed to tackle today’s international problems. Their experience with the subject supports their decision to write the book and not edit a sequence of separately written chapters. As a result, the book also reads easily; the reader does not have to adapt to various writing styles and benefits from a unified reference list and index.

The book cover clearly identifies the volume as one aimed at a professional audience, and most of the topics will indeed prove difficult to assimilate by the general public. However, even a nonprofessional could greatly benefit from reading the first two chapters (“Introduction” and “Subjects Allied to Gait”). Here, the study of gait is presented methodically, starting with early historical studies (when it was associated with psychology and medical sciences) and ending with the current state of the art supported by government-based programs focused on human identification. In addition, early models of gait are presented and discussed. Reading about these topics, one realizes that the concept of gait is not just another fad in the ever-expanding authentication-for-security field, but a topic with a strong past and wide implications to other fields as well. In this view, the next chapter, discussing the current state of the art in gait data collection, also presents an interesting topic, allowing for deeper insight into how contemporary identification methods are tested and evaluated, and hinting at the current limitations.

The chapters contain a list of recent developments in gait-based authentication research, with the main focus seemingly on the directions followed by the authors. While this allows for a deeper understanding of such directions, the reader may find himself or herself deprived of a better understanding of the breadth of the field as exemplified by the considerable list of studies cited throughout the volume. From the point of view of a beginning professional, it would have been interesting to see a more detailed discussion of how the feature data space is formed for the gait analysis, complementing the excellent classification and recognition techniques that were designed to be applied to such a data space. Doing so, the authors would have ended with a comprehensive volume that better fits the profile of one of the first books fully dedicated to the subject. This factor does not decrease the overall value of the text.

An interesting component is the inclusion of documentation related to data collection, such as the lab setup and subject instruction-consent forms used to collect gait data. Given that an interested reader may want to create a new dataset, such documents provide unique insight into the process, and provide valuable lessons based on the authors’ experience.

As the authors mention, compared with other biometrics such as iris or fingerprint recognition, gait-based identification has the advantage of taking measurements from a distance and often using low-resolution data. At the same time, the methods developed thus far are still limited by computational restrictions, and their performance is affected by a long list of factors, such as illumination, changing the viewing angle and/or partial occlusion of the subject, weight carried, clothing, and mood. With these considerations, the future of gait-based identification may be geared toward a more hybrid deployment (as a support to other biometrics) than as a standalone tool. Nevertheless, new and exciting directions are also possible, such as the use of gait-based identification in the surveillance field for identification of “interesting events” and the improvement of human animation. Based on this, a person interested in human-movement modeling and authentication would benefit from reading the book as the first step in understanding both the capabilities and limitations of the field.