While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet.

The ability to accurately identify human activities is essential for developing automatic rehabilitation and sports training systems. In this paper, large-scale exercise motion data obtained from a forearm-worn wearable sensor are classified with a convolutional neural network (CNN). Time-series data consisting of accelerometer and orientation measurements are formatted as images, allowing the CNN to automatically extract discriminative features. A comparative study on the effects of image formatting and different CNN architectures is also presented. The best performing configuration classifies 50 gym exercises with 92.1% accuracy.

In the past decade, model learning techniques have provided appealing approaches for determining the dynamic model of robots from data. These techniques strongly capture the complicated effects of robot dynamics, which are often neglected in hand-crafted dynamic models. However, unlike robust performance shown in trained tasks, learned models do not exhibit a reliable performance in new tasks as they are valid only near the domain of the trained tasks. In this paper, we propose an alternative approach for task-to-task transfer learning, called “Independent Joint Learning (IJL).” IJL learns the model for each joint independently rather than the whole body at one time to effectively transfer knowledge between tasks. A comparative simulation study on a 6 DOF PUMA robot demonstrates that our approach outperforms other related approaches when a task different from trained tasks is proposed.

Motion planning for robots subject to holonomic constraints typically involves planning on constraint manifolds. In this paper we present the Tangent Space Rapidly Exploring Random Tree (TS-RRT) algorithm for planning on constraint manifolds. The key idea is to construct random trees not on the constraint manifold itself, but rather on tangent space approximations to the constraint manifold. Curvature-based methods are developed for constructing bounded tangent space approximations, as well as procedures for random node generation and bidirectional tree extension. Extensive numerical experiments suggest that the TS-RRT algorithm, despite its increased preprocessing and bookkeeping, outperforms existing constrained planning algorithms for a wide range of benchmark planning problems.