Proceedings Article•
A public domain dataset for human activity recognition using smartphones
01 Jan 2013-pp 437-442
TL;DR: An Activity Recognition database is described, built from the recordings of 30 subjects doing Activities of Daily Living while carrying a waist-mounted smartphone with embedded inertial sensors, which is released to public domain on a well-known on-line repository.
Abstract: Human-centered computing is an emerging research field that aims to understand human behavior and integrate users and their social context with computer systems. One of the most recent, challenging and appealing applications in this framework consists in sensing human body motion using smartphones to gather context information about people actions. In this context, we describe in this work an Activity Recognition database, built from the recordings of 30 subjects doing Activities of Daily Living (ADL) while carrying a waist-mounted smartphone with embedded inertial sensors, which is released to public domain on a well-known on-line repository. Results, obtained on the dataset by exploiting a multiclass Support Vector
Machine (SVM), are also acknowledged.
Citations
More filters
Proceedings Article•
02 Nov 2017TL;DR: In this paper, the authors propose a method to learn deep ReLU-based classifiers that are provably robust against norm-bounded adversarial perturbations on the training data; for previously unseen examples, the approach will be guaranteed to detect all adversarial examples, though it may flag some non-adversarial examples as well.
Abstract: We propose a method to learn deep ReLU-based classifiers that are provably robust against norm-bounded adversarial perturbations (on the training data; for previously unseen examples, the approach will be guaranteed to detect all adversarial examples, though it may flag some non-adversarial examples as well). The basic idea of the approach is to consider a convex outer approximation of the set of activations reachable through a norm-bounded perturbation, and we develop a robust optimization procedure that minimizes the worst case loss over this outer region (via a linear program). Crucially, we show that the dual problem to this linear program can be represented itself as a deep network similar to the backpropagation network, leading to very efficient optimization approaches that produce guaranteed bounds on the robust loss. The end result is that by executing a few more forward and backward passes through a slightly modified version of the original network (though possibly with much larger batch sizes), we can learn a classifier that is provably robust to any norm-bounded adversarial attack. We illustrate the approach on a toy 2D robust classification task, and on a simple convolutional architecture applied to MNIST, where we produce a classifier that provably has less than 8.4% test error for any adversarial attack with bounded $\ell_\infty$ norm less than $\epsilon = 0.1$. This represents the largest verified network that we are aware of, and we discuss future challenges in scaling the approach to much larger domains.
934 citations
TL;DR: The concept of federated learning (FL) as mentioned in this paperederated learning has been proposed to enable collaborative training of an ML model and also enable DL for mobile edge network optimization in large-scale and complex mobile edge networks, where heterogeneous devices with varying constraints are involved.
Abstract: In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications, e.g., for medical purposes and in vehicular networks. Traditional cloud-based Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL.
895 citations
Proceedings Article•
04 Dec 2017TL;DR: In this paper, the authors propose a novel systems-aware optimization method, MOCHA, that is robust to practical systems issues, such as high communication cost, stragglers, and fault tolerance for distributed multi-task learning.
Abstract: Federated learning poses new statistical and systems challenges in training machine learning models over distributed networks of devices. In this work, we show that multi-task learning is naturally suited to handle the statistical challenges of this setting, and propose a novel systems-aware optimization method, MOCHA, that is robust to practical systems issues. Our method and theory for the first time consider issues of high communication cost, stragglers, and fault tolerance for distributed multi-task learning. The resulting method achieves significant speedups compared to alternatives in the federated setting, as we demonstrate through simulations on real-world federated datasets.
894 citations
TL;DR: A deep convolutional neural network is proposed to perform efficient and effective HAR using smartphone sensors by exploiting the inherent characteristics of activities and 1D time-series signals, at the same time providing a way to automatically and data-adaptively extract robust features from raw data.
Abstract: This paper proposes a deep convolutional neural network for HAR using smartphone sensors.Experiments show that the proposed method derives relevant and more complex features.The method achieved an almost perfect classification on moving activities.It outperforms other state-of-the-art data mining techniques in HAR. Human activities are inherently translation invariant and hierarchical. Human activity recognition (HAR), a field that has garnered a lot of attention in recent years due to its high demand in various application domains, makes use of time-series sensor data to infer activities. In this paper, a deep convolutional neural network (convnet) is proposed to perform efficient and effective HAR using smartphone sensors by exploiting the inherent characteristics of activities and 1D time-series signals, at the same time providing a way to automatically and data-adaptively extract robust features from raw data. Experiments show that convnets indeed derive relevant and more complex features with every additional layer, although difference of feature complexity level decreases with every additional layer. A wider time span of temporal local correlation can be exploited (1?9-1?14) and a low pooling size (1?2-1?3) is shown to be beneficial. Convnets also achieved an almost perfect classification on moving activities, especially very similar ones which were previously perceived to be very difficult to classify. Lastly, convnets outperform other state-of-the-art data mining techniques in HAR for the benchmark dataset collected from 30 volunteer subjects, achieving an overall performance of 94.79% on the test set with raw sensor data, and 95.75% with additional information of temporal fast Fourier transform of the HAR data set.
854 citations
Posted Content•
TL;DR: A method to learn deep ReLU-based classifiers that are provably robust against norm-bounded adversarial perturbations, and it is shown that the dual problem to this linear program can be represented itself as a deep network similar to the backpropagation network, leading to very efficient optimization approaches that produce guaranteed bounds on the robust loss.
Abstract: We propose a method to learn deep ReLU-based classifiers that are provably robust against norm-bounded adversarial perturbations on the training data. For previously unseen examples, the approach is guaranteed to detect all adversarial examples, though it may flag some non-adversarial examples as well. The basic idea is to consider a convex outer approximation of the set of activations reachable through a norm-bounded perturbation, and we develop a robust optimization procedure that minimizes the worst case loss over this outer region (via a linear program). Crucially, we show that the dual problem to this linear program can be represented itself as a deep network similar to the backpropagation network, leading to very efficient optimization approaches that produce guaranteed bounds on the robust loss. The end result is that by executing a few more forward and backward passes through a slightly modified version of the original network (though possibly with much larger batch sizes), we can learn a classifier that is provably robust to any norm-bounded adversarial attack. We illustrate the approach on a number of tasks to train classifiers with robust adversarial guarantees (e.g. for MNIST, we produce a convolutional classifier that provably has less than 5.8% test error for any adversarial attack with bounded $\ell_\infty$ norm less than $\epsilon = 0.1$), and code for all experiments in the paper is available at this https URL.
749 citations
Cites background from "A public domain dataset for human a..."
...HAR We present results on a human activity recognition dataset (Anguita et al., 2013)....
[...]
References
More filters
TL;DR: High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated and the performance of the support- vector network is compared to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.
Abstract: The support-vector network is a new learning machine for two-group classification problems. The machine conceptually implements the following idea: input vectors are non-linearly mapped to a very high-dimension feature space. In this feature space a linear decision surface is constructed. Special properties of the decision surface ensures high generalization ability of the learning machine. The idea behind the support-vector network was previously implemented for the restricted case where the training data can be separated without errors. We here extend this result to non-separable training data.
High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated. We also compare the performance of the support-vector network to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.
37,861 citations
"A public domain dataset for human a..." refers methods in this paper
...The classification results using the multiclass SVM (MC-SVM) for the 6 ADL are presented in Table 4....
[...]
...For this purpose, we exploit well-known and state-of-the-art Support Vector Machine (SVM) [13] binary classifiers, which are generalized to the multiclass case through a One-Vs-All (OVA) approach: the SVM hyperparameters are selected through a 10-fold Cross Validation procedure and Gaussian kernels are used for our experiments....
[...]
...Some results, obtained by exploiting a multi class Support Vector Machine (SVM) classifier [13], are shown as well....
[...]
...Results, obtained on the dataset by exploiting a multiclass Support Vector Machine (SVM), are also acknowledged....
[...]
...It is also worth underlining that the MC-SVM model outperforms by 7% the classifier learned on our previous dataset described in [19], where only acceleration data from the smartphone were taken into account for the recognition: this suggests that the new features, introduced in the publicly available dataset as depicted in Section 2.2, allow to ease the learning process....
[...]
21 Apr 2004
TL;DR: This is the first work to investigate performance of recognition algorithms with multiple, wire-free accelerometers on 20 activities using datasets annotated by the subjects themselves, and suggests that multiple accelerometers aid in recognition.
Abstract: In this work, algorithms are developed and evaluated to de- tect physical activities from data acquired using five small biaxial ac- celerometers worn simultaneously on different parts of the body. Ac- celeration data was collected from 20 subjects without researcher su- pervision or observation. Subjects were asked to perform a sequence of everyday tasks but not told specifically where or how to do them. Mean, energy, frequency-domain entropy, and correlation of acceleration data was calculated and several classifiers using these features were tested. De- cision tree classifiers showed the best performance recognizing everyday activities with an overall accuracy rate of 84%. The results show that although some activities are recognized well with subject-independent training data, others appear to require subject-specific training data. The results suggest that multiple accelerometers aid in recognition because conjunctions in acceleration feature values can effectively discriminate many activities. With just two biaxial accelerometers - thigh and wrist - the recognition performance dropped only slightly. This is the first work to investigate performance of recognition algorithms with multiple, wire-free accelerometers on 20 activities using datasets annotated by the subjects themselves.
3,223 citations
Additional excerpts
...sensor repositioning after dressing [5])....
[...]
TL;DR: This work describes and evaluates a system that uses phone-based accelerometers to perform activity recognition, a task which involves identifying the physical activity a user is performing, and has a wide range of applications, including automatic customization of the mobile device's behavior based upon a user's activity.
Abstract: Mobile devices are becoming increasingly sophisticated and the latest generation of smart cell phones now incorporates many diverse and powerful sensors These sensors include GPS sensors, vision sensors (ie, cameras), audio sensors (ie, microphones), light sensors, temperature sensors, direction sensors (ie, magnetic compasses), and acceleration sensors (ie, accelerometers) The availability of these sensors in mass-marketed communication devices creates exciting new opportunities for data mining and data mining applications In this paper we describe and evaluate a system that uses phone-based accelerometers to perform activity recognition, a task which involves identifying the physical activity a user is performing To implement our system we collected labeled accelerometer data from twenty-nine users as they performed daily activities such as walking, jogging, climbing stairs, sitting, and standing, and then aggregated this time series data into examples that summarize the user activity over 10- second intervals We then used the resulting training data to induce a predictive model for activity recognition This work is significant because the activity recognition model permits us to gain useful knowledge about the habits of millions of users passively---just by having them carry cell phones in their pockets Our work has a wide range of applications, including automatic customization of the mobile device's behavior based upon a user's activity (eg, sending calls directly to voicemail if a user is jogging) and generating a daily/weekly activity profile to determine if a user (perhaps an obese child) is performing a healthy amount of exercise
2,417 citations
"A public domain dataset for human a..." refers background in this paper
...Consequently, in the last few years, some works aiming to understand human behavior using smartphones have been proposed: for instance in [6], one of the first approaches to exploit an Android smartphone for HAR employing its embedded triaxial accelerometers; additional results have also been presented in [7, 8]....
[...]
Proceedings Article•
09 Jul 2005TL;DR: This paper reports on the efforts to recognize user activity from accelerometer data and performance of base-level and meta-level classifiers, and Plurality Voting is found to perform consistently well across different settings.
Abstract: Activity recognition fits within the bigger framework of context awareness. In this paper, we report on our efforts to recognize user activity from accelerometer data. Activity recognition is formulated as a classification problem. Performance of base-level classifiers and meta-level classifiers is compared. Plurality Voting is found to perform consistently well across different settings.
1,561 citations
"A public domain dataset for human a..." refers background in this paper
...Some approaches have adapted dedicated motion sensors in different body parts such as the waist, wrist, chest and thighs achieving good classification performance [4]....
[...]