scispace - formally typeset
Search or ask a question
Author

Nils Y. Hammerla

Bio: Nils Y. Hammerla is an academic researcher from Newcastle University. The author has contributed to research in topics: Activity recognition & Feature extraction. The author has an hindex of 25, co-authored 41 publications receiving 4711 citations.

Papers
More filters
Posted Content
TL;DR: A novel attention gate (AG) model for medical imaging that automatically learns to focus on target structures of varying shapes and sizes is proposed to eliminate the necessity of using explicit external tissue/organ localisation modules of cascaded convolutional neural networks (CNNs).
Abstract: We propose a novel attention gate (AG) model for medical imaging that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using explicit external tissue/organ localisation modules of cascaded convolutional neural networks (CNNs). AGs can be easily integrated into standard CNN architectures such as the U-Net model with minimal computational overhead while increasing the model sensitivity and prediction accuracy. The proposed Attention U-Net architecture is evaluated on two large CT abdominal datasets for multi-class image segmentation. Experimental results show that AGs consistently improve the prediction performance of U-Net across different datasets and training sizes while preserving computational efficiency. The code for the proposed architecture is publicly available.

2,452 citations

Journal ArticleDOI
01 Feb 2017-PLOS ONE
TL;DR: The summary measure of overall physical activity is lower in older participants and age-related differences in activity are most prominent in the afternoon and evening, which lays the foundation for studies of physical activity and its health consequences.
Abstract: BACKGROUND: Physical activity has not been objectively measured in prospective cohorts with sufficiently large numbers to reliably detect associations with multiple health outcomes. Technological advances now make this possible. We describe the methods used to collect and analyse accelerometer measured physical activity in over 100,000 participants of the UK Biobank study, and report variation by age, sex, day, time of day, and season. METHODS: Participants were approached by email to wear a wrist-worn accelerometer for seven days that was posted to them. Physical activity information was extracted from 100Hz raw triaxial acceleration data after calibration, removal of gravity and sensor noise, and identification of wear / non-wear episodes. We report age- and sex-specific wear-time compliance and accelerometer measured physical activity, overall and by hour-of-day, week-weekend day and season. RESULTS: 103,712 datasets were received (44.8% response), with a median wear-time of 6.9 days (IQR:6.5-7.0). 96,600 participants (93.3%) provided valid data for physical activity analyses. Vector magnitude, a proxy for overall physical activity, was 7.5% (2.35mg) lower per decade of age (Cohen's d = 0.9). Women had a higher vector magnitude than men, apart from those aged 45-54yrs. There were major differences in vector magnitude by time of day (d = 0.66). Vector magnitude differences between week and weekend days (d = 0.12 for men, d = 0.09 for women) and between seasons (d = 0.27 for men, d = 0.15 for women) were small. CONCLUSIONS: It is feasible to collect and analyse objective physical activity data in large studies. The summary measure of overall physical activity is lower in older participants and age-related differences in activity are most prominent in the afternoon and evening. This work lays the foundation for studies of physical activity and its health consequences. Our summary variables are part of the UK Biobank dataset and can be used by researchers as exposures, confounding factors or outcome variables in future analyses.

653 citations

Proceedings Article
13 Feb 2017
TL;DR: The authors showed that the linear transformation between two spaces should be orthogonal and proposed a novel "inverted softmax" for identifying translation pairs, with which they improved the precision of Mikolov's original mapping from 34% to 43%.
Abstract: Usually bilingual word vectors are trained "online". Mikolov et al. showed they can also be found "offline", whereby two pre-trained embeddings are aligned with a linear transformation, using dictionaries compiled from expert knowledge. In this work, we prove that the linear transformation between two spaces should be orthogonal. This transformation can be obtained using the singular value decomposition. We introduce a novel "inverted softmax" for identifying translation pairs, with which we improve the precision @1 of Mikolov's original mapping from 34% to 43%, when translating a test set composed of both common and rare English words into Italian. Orthogonal transformations are more robust to noise, enabling us to learn the transformation without expert bilingual signal by constructing a "pseudo-dictionary" from the identical character strings which appear in both languages, achieving 40% precision on the same test set. Finally, we extend our method to retrieve the true translations of English sentences from a corpus of 200k Italian sentences with a precision @1 of 68%.

492 citations

Proceedings Article
09 Jul 2016
TL;DR: This paper rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors, and describes how to train recurrent approaches in this setting, introduces a novel regularisation approach, and illustrates how they outperform the state-of-the-art on a large benchmark dataset.
Abstract: Human activity recognition (HAR) in ubiquitous computing is beginning to adopt deep learning to substitute for well-established analysis techniques that rely on hand-crafted feature extraction and classification methods. However, from these isolated applications of custom deep architectures it is difficult to gain an overview of their suitability for problems ranging from the recognition of manipulative gestures to the segmentation and identification of physical activities like running or ascending stairs. In this paper we rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors. We describe how to train recurrent approaches in this setting, introduce a novel regularisation approach, and illustrate how they outperform the state-of-the-art on a large benchmark dataset. We investigate the suitability of each model for HAR, across thousands of recognition experiments with randomly sampled model configurations, explore the impact of hyperparameters using the fANOVA framework, and provide guidelines for the practitioner who wants to apply deep learning in their problem setting.

463 citations

Posted Content
TL;DR: In this paper, the authors rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors, and illustrate how they outperform the state-of-the-art on a large benchmark dataset.
Abstract: Human activity recognition (HAR) in ubiquitous computing is beginning to adopt deep learning to substitute for well-established analysis techniques that rely on hand-crafted feature extraction and classification techniques. From these isolated applications of custom deep architectures it is, however, difficult to gain an overview of their suitability for problems ranging from the recognition of manipulative gestures to the segmentation and identification of physical activities like running or ascending stairs. In this paper we rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors. We describe how to train recurrent approaches in this setting, introduce a novel regularisation approach, and illustrate how they outperform the state-of-the-art on a large benchmark dataset. Across thousands of recognition experiments with randomly sampled model configurations we investigate the suitability of each model for different tasks in HAR, explore the impact of hyperparameters using the fANOVA framework, and provide guidelines for the practitioner who wants to apply deep learning in their problem setting.

390 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis.
Abstract: Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

13,246 citations

Journal ArticleDOI
11 Oct 2018-Nature
TL;DR: Deep phenotype and genome-wide genetic data from 500,000 individuals from the UK Biobank is described, describing population structure and relatedness in the cohort, and imputation to increase the number of testable variants to 96 million.
Abstract: The UK Biobank project is a prospective cohort study with deep genetic and phenotypic data collected on approximately 500,000 individuals from across the United Kingdom, aged between 40 and 69 at recruitment. The open resource is unique in its size and scope. A rich variety of phenotypic and health-related information is available on each participant, including biological measurements, lifestyle indicators, biomarkers in blood and urine, and imaging of the body and brain. Follow-up information is provided by linking health and medical records. Genome-wide genotype data have been collected on all participants, providing many opportunities for the discovery of new genetic associations and the genetic bases of complex traits. Here we describe the centralized analysis of the genetic data, including genotype quality, properties of population structure and relatedness of the genetic data, and efficient phasing and genotype imputation that increases the number of testable variants to around 96 million. Classical allelic variation at 11 human leukocyte antigen genes was imputed, resulting in the recovery of signals with known associations between human leukocyte antigen alleles and many diseases.

4,489 citations

21 Jun 2010

1,966 citations

Journal ArticleDOI
18 Jan 2016-Sensors
TL;DR: A generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which is suitable for multimodal wearable sensors, does not require expert knowledge in designing features, and explicitly models the temporal dynamics of feature activations is proposed.
Abstract: Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation.

1,896 citations

Journal ArticleDOI
TL;DR: The recent advance of deep learning based sensor-based activity recognition is surveyed from three aspects: sensor modality, deep model, and application and detailed insights on existing work are presented and grand challenges for future research are proposed.

1,334 citations