Evaluation of supervised classification algorithms for human activity recognition with inertial sensors

Zebin, Tahmina; Scully, Patricia; Ozanyan, Krikor B.

doi:10.1109/icsens.2017.8234222

Cited by 28 publications

(16 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To place our results in context Table I summaries the performance of other classification techniques when applied to the same data set as used in this paper, and to previous LSTM models. Our proposed LSTM+BN that processes featureless raw signals achieves 92% overall classification accuracy which is slightly lower than the Support Vector Machine (SVM) method in [14] which used handcrafted features to achieve 93.4% accuracy. This compromise in accuracy can be discounted by the fact that the LSTM classification is more generalized and capable detecting activities that have long term dependence which is not the case for SVM.…”

Section: Comparison With Other Approachesmentioning

confidence: 90%

“…To evaluate the performance of the model, we processed the time series data from a waist mounted inertial sensor recorded at 50 Hz sampling frequency containing both accelerometer and gyroscope measurements. Data for 20 subjects is present, described in detail in [14]. The dataset contains six everyday activities: 0-walk on level surface; 1-walk upstairs; 2-walk downstairs; 3-sitting; 4-standing; 5-lying.…”

Section: B Dataset Preparationmentioning

confidence: 99%

See 1 more Smart Citation

Human activity recognition from inertial sensor time-series using batch normalized deep LSTM recurrent networks

Zebin

Sperrin

Peek

et al. 2018

2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

Self Cite

View full text Add to dashboard Cite

In recent years machine learning methods for human activity recognition have been found very effective. These classify discriminative features generated from raw input sequences acquired from body-worn inertial sensors. However, it involves an explicit feature extraction stage from the raw data, and although human movements are encoded in a sequence of successive samples in time most state-of-theart machine learning methods do not exploit the temporal correlations between input data samples. In this paper we present a Long-Short Term Memory (LSTM) deep recurrent neural network for the classification of six daily life activities from accelerometer and gyroscope data. Results show that our LSTM can processes featureless raw input signals, and achieves 92% average accuracy in a multi-class-scenario. Further, we show that this accuracy can be achieved with almost four times fewer training epochs by using a batch normalization approach.

show abstract

Section: Comparison With Other Approachesmentioning

confidence: 90%

Section: B Dataset Preparationmentioning

confidence: 99%

Human activity recognition from inertial sensor time-series using batch normalized deep LSTM recurrent networks

Zebin

Sperrin

Peek

et al. 2018

2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Rodriguez et al [23] reported BioHarness and smartphone based activity recognition using decision tree for the classification. Zebin et al [24] conducted the comparison between different HAR algorithms for the inertial sensor data. Tang and Sazonov [9] compared the ANN and SVM classification algorithms for the smart shoe.…”

Section: Related Workmentioning

confidence: 99%

Efficient FPGA Implementation of Multilayer Perceptron for Real-Time Human Activity Classification

et al. 2019

View full text Add to dashboard Cite

The smartphone-based human activity recognition (HAR) systems are not capable to deliver high-end performance for challenging applications. We propose a dedicated hardware-based HAR system for smart military wearables, which uses a multilayer perceptron (MLP) algorithm to perform activity classification. To achieve the flexible and efficient hardware design, the inherent MLP architecture with parallel computation is implemented on FPGA. The system performance has been evaluated using the UCI human activity dataset with 7767 feature samples of 20 subjects. The three combinations of a dataset are trained, validated, and tested on ten different MLP models with distinct topologies. The MLP design with the 7-6-5 topology is finalized from the classification accuracy and cross entropy performance. The five versions of the final MLP design (7-6-5) with different data precision are implemented on FPGA. The analysis shows that the MLP designed with 16-bit fixed-point data precision is the most efficient MLP implementation in the context of classification accuracy, resource utilization, and power consumption. The proposed MLP design requires only 270 ns for classification and consumes 120 mW of power. The recognition accuracy and hardware results performance achieved are better than many of the recently reported works. INDEX TERMS Human activity recognition, multilayer perceptron, smart military wearables, classifier hardware implementation, field programmable gate array.

show abstract

“…We analyse two datasets, henceforth Dataset 1 and Dataset 2. Dataset 1 analysed in [19], contained 661056 observations from 20 participants on activity labels collected at 50 Hz sampling frequency. Six labelled activities were encoded: 0-walk on level surface; 1-walk upstairs; 2-walk downstairs; 3-sitting; 4-standing; 5-lying.…”

Section: A Data Collectionmentioning

confidence: 99%

Downsampling wearable sensor data packets by measuring their information value

2019

View full text Add to dashboard Cite

Long-Short Term Memory models (LSTMs) are data-driven routines that classify Human Activity Recognition (HAR) with minimum human input. The price to pay for analysing large sequences of on-body sensor measurements with LSTMs are high processing power and battery requirements. In this paper, we recognize that sensor data packets have differing information value to classify HAR and propose to quantify it with cross entropy (CrossEn), Kullback Leibler (KL) divergence and sample entropy (SampEn). Both, CrossEn and SampEn have the potential to guide dropping redundant data packets without compromising HAR. However, we do not find substantial improvements in dropping rates when downsampling by CrossEn and SampEn over computationally cheaper random and uniform alternatives. Our results show that the KL divergence, evaluated at training time is equivalent to the classification accuracy criteria that involves a testing set. The computational requirements to compute the KL in real-time could well guide sensor node design to downsample wearable measurements near the user.

show abstract

Evaluation of supervised classification algorithms for human activity recognition with inertial sensors

Cited by 28 publications

References 13 publications

Human activity recognition from inertial sensor time-series using batch normalized deep LSTM recurrent networks

Human activity recognition from inertial sensor time-series using batch normalized deep LSTM recurrent networks

Efficient FPGA Implementation of Multilayer Perceptron for Real-Time Human Activity Classification

Downsampling wearable sensor data packets by measuring their information value

Contact Info

Product

Resources

About