Biometric recognition through gait analysis

Abstract

The use of people recognition techniques has become critical in some areas. For instance, social or assistive robots carry out collaborative tasks in the robotics field. A robot must know who to work with to deal with such tasks. Using biometric patterns may replace identification cards or codes on access control to critical infrastructures. The usage of Red Green Blue Depth (RGBD) cameras is ubiquitous to solve people recognition. However, this sensor has some constraints, such as they demand high computational capabilities, require the users to face the sensor, or do not regard users’ privacy. Furthermore, in the COVID-19 pandemic, masks hide a significant portion of the face. In this work, we present BRITTANY, a biometric recognition tool through gait analysis using Laser Imaging Detection and Ranging (LIDAR) data and a Convolutional Neural Network (CNN). A Proof of Concept (PoC) has been carried out in an indoor environment with five users to evaluate BRITTANY. A new CNN architecture is presented, allowing the classification of aggregated occupancy maps that represent the people’s gait. This new architecture has been compared with LeNet-5 and AlexNet through the same datasets. The final system reports an accuracy of 88%.

Ensemble of Heterogeneous Base Classifiers for Human Gait Recognition

“…A drawback of such systems is that they are susceptible to activities allowing resource access to people who are unauthorized to do so. Dynamic systems based on measuring human behaviors, on the other hand, such as voice [ 6 ], handwriting [ 7 ], keystroke dynamics [ 8 ] or gait [ 9 , 10 ] recognition, although characterized by lower accuracy, are either very difficult or impossible to deceive. This is the reason that multimodal biometric systems using two or more biometric solutions of both types are often recommended [ 11 , 12 ].…”

Gait signature changes with walking speed are similar among able-bodied young adults despite persistent individual-specific differences

“…Our work can potentially aid researchers with determining the minimum equipment required for constructing gait signatures capable of effectively characterizing individual differences in gait dynamics. The ability of gait signatures to reliably identify individuals walking at any gait speed may also be useful in research where biometric recognition using machine learning is important [76,77] . The finding that 3D MDS gait signatures can be reasonably predicted from a limited set of speeds may offer a valuable framework similar to velocity-based training in sports conditioning [78,79] , potentially enabling trainers to pinpoint optimal training speeds for athletes or prescribe training intensities to induce desired changes in in movement quality.…”

This Was (Not) Intended: How Intent Communication and Biometrics Can Enhance Social Interactions With Robots

“…Álvarez-Aparicio used the Laser Imaging Detection and Ranging (LIDAR) sensor on a non-humanoid robot to identify individuals based on their gait [2], reaching a precision score of 88%. Although their approach was not specifically designed for SARs, it illustrates a possible use case where a robot identifies a user walking from a distance, understands their preferences, and acts accordingly.…”