Abstract. We propose a novel method for vision based simultaneous localization and mapping (vSLAM) using a biologically inspired vision sensor that mimics the human retina. The sensor consists of a 128x128 array of asynchronously operating pixels, which independently emit events upon a temporal illumination change. Such a representation generates small amounts of data with high temporal precision; however, most classic computer vision algorithms need to be reworked as they require full RGB(-D) images at fixed frame rates. Our presented vSLAM algorithm operates on individual pixel events and generates high-quality 2D environmental maps with precise robot localizations. We evaluate our method with a state-of-the-art marker-based external tracking system and demonstrate real-time performance on standard computing hardware.

Gait patterns are a result of the complex kinematics that enable human two-legged locomotion, and they can reveal a lot about a person’s state and health. Analysing them is useful for researchers to get new insights into the course of diseases, and for physicians to track the progress after healing from injuries. When a person walks and is interfered with in any way, the resulting disturbance can show up and be found in the gait patterns. This paper describes an experimental setup for capturing gait patterns with a capacitive sensor floor, which can detect the time and position of foot contacts on the floor. With this setup, a dataset was recorded where 42 participants walked over a sensor floor in different modes, inter alia, normal pace, closed eyes, and dual-task. A recurrent neural network based on Long Short-Term Memory units was trained and evaluated for the classification task of recognising the walking mode solely from the floor sensor data. Furthermore, participants were asked to do the Unilateral Heel-Rise Test, and their gait was recorded before and after doing the test. Another neural network instance was trained to predict the number of repetitions participants were able to do on the test. As the results of the classification tasks turned out to be promising, the combination of this sensor floor and the recurrent neural network architecture seems like a good system for further investigation leading to applications in health and care.

No abstract

Abstract-In this paper we present an autonomous mobile robot setting that automatically explores and maps unknown indoor environments, exclusively with information from an embedded event-based dynamic vision sensor (eDVS) and a ring of bump switches on the robot. The eDVS provides a sparse pre-processed visual signature of the currently visible patch of ceiling, which is used for real-time simultaneous localization and mapping (SLAM). Signals from the robot's bump switches together with its current position estimate continuously improve the system's reasoning about traversable areas. A heuristic path planning method motivated by the A * search algorithm generates routes for continuous autonomous exploration. We demonstrate robust real-time operation and evaluate the performance of our system in various indoor environments.

No abstract