Recently, deep learning (DL) methods have been introduced very successfully into human activity recognition (HAR) scenarios in ubiquitous and wearable computing. Especially the prospect of overcoming the need for manual feature design combined with superior classification capabilities render deep neural networks very attractive for real-life HAR applications. Even though DL-based approaches now outperform the state-of-the-art in a number of recognition tasks, still substantial challenges remain. Most prominently, issues with real-life datasets, typically including imbalanced datasets and problematic data quality, still limit the effectiveness of activity recognition using wearables. In this paper we tackle such challenges through Ensembles of deep Long Short Term Memory (LSTM) networks. LSTM networks currently represent the state-of-the-art with superior classification performance on relevant HAR benchmark datasets. We have developed modified training procedures for LSTM networks and combine sets of diverse LSTM learners into classifier collectives. We demonstrate that Ensembles of deep LSTM learners outperform individual LSTM networks and thus push the state-of-the-art in human activity recognition using wearables. Through an extensive experimental evaluation on three standard benchmarks (Opportunity, PAMAP2, Skoda) we demonstrate the excellent recognition capabilities of our approach and its potential for real-life applications of human activity recognition.

In recent years, there has been a significant expansion in the development and use of multi-modal sensors and technologies to monitor physical activity, sleep and circadian rhythms. These developments make accurate sleep monitoring at scale a possibility for the first time. Vast amounts of multi-sensor data are being generated with potential applications ranging from large-scale epidemiological research linking sleep patterns to disease, to wellness applications, including the sleep coaching of individuals with chronic conditions. However, in order to realise the full potential of these technologies for individuals, medicine and research, several significant challenges must be overcome. There are important outstanding questions regarding performance evaluation, as well as data storage, curation, processing, integration, modelling and interpretation. Here, we leverage expertise across neuroscience, clinical medicine, bioengineering, electrical engineering, epidemiology, computer science, mHealth and human–computer interaction to discuss the digitisation of sleep from a inter-disciplinary perspective. We introduce the state-of-the-art in sleep-monitoring technologies, and discuss the opportunities and challenges from data acquisition to the eventual application of insights in clinical and consumer settings. Further, we explore the strengths and limitations of current and emerging sensing methods with a particular focus on novel data-driven technologies, such as Artificial Intelligence.

Recognising remote sensing scene images remains challenging due to large visual-semantic discrepancies. These mainly arise due to the lack of detailed annotations that can be employed to align pixel-level representations with high-level semantic labels. As the tagging process is labour-intensive and subjective, we hereby propose a novel Multi-Granularity Canonical Appearance Pooling (MG-CAP) to automatically capture the latent ontological structure of remote sensing datasets. We design a granular framework that allows progressively cropping the input image to learn multi-grained features. For each specific granularity, we discover the canonical appearance from a set of pre-defined transformations and learn the corresponding CNN features through a maxout-based Siamese style architecture. Then, we replace the standard CNN features with Gaussian covariance matrices and adopt the proper matrix normalisations for improving the discriminative power of features. Besides, we provide a stable solution for training the eigenvalue-decomposition function (EIG) in a GPU and demonstrate the corresponding back-propagation using matrix calculus. Extensive experiments have shown that our framework can achieve promising results in public remote sensing scene datasets.

Parkinson’s disease (PD) is the second most common neurodegenerative disease; gait impairments are typical and are associated with increased fall risk and poor quality of life. Gait is potentially a useful biomarker to help discriminate PD at an early stage, however the optimal characteristics and combination are unclear. In this study, we used machine learning (ML) techniques to determine the optimal combination of gait characteristics to discriminate people with PD and healthy controls (HC). 303 participants (119 PD, 184 HC) walked continuously around a circuit for 2-minutes at a self-paced walk. Gait was quantified using an instrumented mat (GAITRite) from which 16 gait characteristics were derived and assessed. Gait characteristics were selected using different ML approaches to determine the optimal method (random forest with information gain and recursive features elimination (RFE) technique with support vector machine (SVM) and logistic regression). Five clinical gait characteristics were identified with RFE-SVM (mean step velocity, mean step length, step length variability, mean step width, and step width variability) that accurately classified PD. Model accuracy for classification of early PD ranged between 73–97% with 63–100% sensitivity and 79–94% specificity. In conclusion, we identified a subset of gait characteristics for accurate early classification of PD. These findings pave the way for a better understanding of the utility of ML techniques to support informed clinical decision-making.

Robust human gait recognition is challenging because of the presence of covariate factors such as carrying condition, clothing, walking surface, etc. In this paper, we model the effect of covariates as an unknown partial feature corruption problem. Since the locations of corruptions may differ for different query gaits, relevant features may become irrelevant when walking condition changes. In this case, it is difficult to train one fixed classifier that is robust to a large number of different covariates. To tackle this problem, we propose a classifier ensemble method based on the random subspace Method (RSM) and majority voting (MV). Its theoretical basis suggests it is insensitive to locations of corrupted features, and thus can generalize well to a large number of covariates. We also extend this method by proposing two strategies, i.e, local enhancing (LE) and hybrid decision-level fusion (HDF) to suppress the ratio of false votes to true votes (before MV). The performance of our approach is competitive against the most challenging covariates like clothing, walking surface, and elapsed time. We evaluate our method on the USF dataset and OU-ISIR-B dataset, and it has much higher performance than other state-of-the-art algorithms.

Unseen Action Recognition (UAR) aims to recognise novel action categories without training examples. While previous methods focus on inner-dataset seen/unseen splits, this paper proposes a pipeline using a large-scale training source to achieve a Universal Representation (UR) that can generalise to a more realistic Cross-Dataset UAR (CDUAR) scenario. We first address UAR as a Generalised Multiple-Instance Learning (GMIL) problem and discover 'building-blocks' from the large-scale ActivityNet dataset using distribution kernels. Essential visual and semantic components are preserved in a shared space to achieve the UR that can efficiently generalise to new datasets. Predicted UR exemplars can be improved by a simple semantic adaptation, and then an unseen action can be directly recognised using UR during the test. Without further training, extensive experiments manifest significant improvements over the UCF101 and HMDB51 benchmarks.