Benchmarking classification techniques using the Opportunity human activity dataset

Sagha, Hesam; Digumarti, Sundara Tejaswi; Millán, José del R.; Chavarriaga, Ricardo; Calatroni, Alberto; Roggen, Daniel; Tröster, Gerhard

doi:10.1109/icsmc.2011.6083628

Cited by 90 publications

(71 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…random forests profit from the several trained trees voting to reach a common verdict. These results and explanations are in line with those reported by Sagha et al [13] for indoor activity recognition. Table 5: F-scores when using different classifiers.…”

Section: Evaluating Different Classifierssupporting

confidence: 82%

See 1 more Smart Citation

Towards Indoor Transportation Mode Detection Using Mobile Sensing

Prentow

Blunck

Kjærgaard

et al. 2015

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

Abstract. Transportation mode detection is a growing field of research, in which a variety of methods have been developed for detecting transportation modes foremost for outdoor travels. It has been employed in application areas such as public transportation, environmental footprint profiling, and context-aware mobile assistants. For indoor travels the problem of transportation mode detection has received comparatively little attention, even though diverse transportation modes, such as biking, electric vehicles, and scooters, are used indoors, especially in large building complexes. The potential applications are diverse, and may also extend beyond indoor variants of the above outdoor applications, and include, e.g., scheduling and progress tracking for mobile workers, management of vehicular resources, and navigation support. However, for indoor transportation mode detection, the physical environment as well as the availability and reliability of sensing resources differ drastically from outdoor scenarios. Owing to these differences, many of the methods developed for outdoor transportation mode detection cannot be easily and reliably applied indoors. In this paper, we explore indoor transportation scenarios to arrive at a conceptual model of indoor transportation modes, and then compare challenges for outdoor and indoor transportation mode detection. In addition, we explore methods for transportation mode detection we deem suitable in indoor settings, and we perform an extensive real-world evaluation of (combinations of) such methods at a large hospital complex. The evaluation presented here utilizes Wi-Fi and accelerometer data collected through smartphones carried by several hospital workers throughout four days of work routines. The results show that the methods can distinguish between six common modes of transportation used by the hospital workers with an F-score of 84.2%.

show abstract

Section: Evaluating Different Classifierssupporting

confidence: 82%

“…For the evaluation, in line with standard practices in activity recognition research, we use the F 1 -score, which is the harmonic mean of precision and recall [13].…”

Section: Evaluation Methodologymentioning

confidence: 99%

Towards Indoor Transportation Mode Detection Using Mobile Sensing

Prentow

Blunck

Kjærgaard

et al. 2015

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

show abstract

“…These sensors include five XSense inertial measurement units (accelerometer, gyro and magnetic sensors) mounted on a custom-made motion jacket, 12 Bluetooth 3-axis acceleration sensors on the limbs and commercial InertiaCube3 inertial sensors located on each foot ( Figure 1b). The challenge ran from May to September 2011 and the outcome was initially announced during a workshop at the IEEE conference on Systems, Man and Cybernetics in Anchorage, Alaska (Sagha et al, 2011a).…”

Section: Challenge On Robust Activity Recognitionmentioning

confidence: 99%

“…A preliminary version of this work was presented at the IEEE conference on Systems, Man and Cybernetics in Anchorage, Alaska (Sagha et al, 2011a). This work has been supported by the EU Future and Emerging Technologies (FET) contract number FP7-Opportunity-225938.…”

Section: Acknowledgementsmentioning

confidence: 99%

The Opportunity challenge: A benchmark database for on-body sensor-based activity recognition

Chavarriaga

Sagha

Calatroni

et al. 2013

Pattern Recognition Letters

Self Cite

603

347

View full text Add to dashboard Cite

There is a growing interest on using ambient and wearable sensors for human activity recognition, fostered by several application domains and wider availability of sensing technologies. This has triggered increasing attention on the development of robust machine learning techniques that exploits multimodal sensor setups. However, unlike other applications, there are no established benchmarking problems for this field. As a matter of fact, methods are usually tested on custom datasets acquired in very specific experimental setups. Furthermore, data is seldom shared between different groups. Our goal is to address this issue by introducing a versatile human activity dataset recorded in a sensor-rich environment. This database was the basis of an open challenge on activity recognition. We report here the outcome of this challenge, as well as baseline performance using different classification techniques. We expect this benchmarking database will motivate other researchers to replicate and outperform the presented results, thus contributing to further advances in the state-of-the-art of activity recognition methods.

show abstract

“…Thus, the miniaturisation and cost reduction of sensor hardware and mobile devices has led to the emergence of research into mobile AR [1]. In several existing studies wearable sensors are used by people while performing their daily activities [1,6], while others additionally use sensors embedded into tools and utensils in an apartment, which allows the analysis of more fine grained activities [8].…”

Section: Introductionmentioning

confidence: 99%

Mobile Activity Recognition Using Ubiquitous Data Stream Mining

João

Gomes

Krishnaswamy

et al. 2012

Data Warehousing and Knowledge Discovery

View full text Add to dashboard Cite

Abstract. Mobile activity recognition focuses on inferring the current activities of a mobile user by leveraging the rich sensory data that is available on today's smart phones and other wearable sensors. The state of the art in mobile activity recognition research has focused on traditional classification learning techniques. In this paper, we propose the Mobile Activity Recognition System (MARS) where for the first time the classifier is built on-board the mobile device itself through ubiquitous data stream mining in an incremental manner. The advantages of on-board data stream mining for mobile activity recognition are: i) personalisation of models built to individual users; ii) increased privacy as the data is not sent to an external site; iii) adaptation of the model as the user's activity profile changes. In our extensive experimental results using a recent benchmarking activity recognition dataset, we show that MARS can achieve similar accuracy when compared with traditional classifiers for activity recognition, while at the same time being scalable and efficient in terms of the mobile device resources consumption. MARS has been implemented on the Android platform for empirical evaluation.

show abstract

Benchmarking classification techniques using the Opportunity human activity dataset

Cited by 90 publications

References 15 publications

Towards Indoor Transportation Mode Detection Using Mobile Sensing

Towards Indoor Transportation Mode Detection Using Mobile Sensing

The Opportunity challenge: A benchmark database for on-body sensor-based activity recognition

Mobile Activity Recognition Using Ubiquitous Data Stream Mining

Contact Info

Product

Resources

About