AUTO-HAR: An adaptive human activity recognition framework using an automated CNN architecture design

Abstract

Convolutional neural networks (CNNs) have demonstrated exceptional results in the analysis of time- series data when used for Human Activity Recognition (HAR). The manual design of such neural architectures is an error-prone and time-consuming process. The search for optimal CNN architectures is considered a revolution in the design of neural networks. By means of Neural Architecture Search (NAS), network architectures can be designed and optimized automatically. Thus, the optimal CNN architecture representation can be found automatically because of its ability to overcome the limitations of human experience and thinking modes. Evolution algorithms, which are derived from evolutionary mechanisms such as natural selection and genetics, have been widely employed to develop and optimize NAS because they can handle a blackbox optimization process for designing appropriate solution representations and search paradigms without explicit mathematical formulations or gradient information. The Genetic optimization algorithm (GA) is widely used to find optimal or near-optimal solutions for difficult problems. Considering these characteristics, an efficient human activity recognition architecture (AUTO-HAR) is presented in this study. Using the evolutionary GA to select the optimal CNN architecture, the current study proposes a novel encoding schema structure and a novel search space with a much broader range of operations to effectively search for the best architectures for HAR tasks. In addition, the proposed search space provides a reasonable degree of depth because it does not limit the maximum length of the devised task architecture. To test the effectiveness of the proposed framework for HAR tasks, three datasets were utilized: UCI-HAR, Opportunity, and DAPHNET. Based on the results of this study, it has been found that the proposed method can efficiently recognize human activity with an average accuracy of 98.5% (∓1.1), 98.3%, and 99.14% (∓0.8) for UCI-HAR, Opportunity, and DAPHNET, respectively.

Keywords: CNN topology; Convolution neural networks; Deep learning; Evolutionary neural network search; Genetic algorithms; Human activity recognition.

Figure 1

The Proposed AUTO-HAR Deep Learning Architecture.

Figure 2

Basic CNN Architecture.

Figure 3

AUTO-CNN architecture design and optimization.

Figure 4

Addition of a basic block to a fixed FC part.

Figure 5

Example of encoding CNN architecture with a depth of two.

Algorithm 1

An algorithm to search for the optimal the CNN architecture.

Algorithm 2

Function: Make POPULATION.

Figure 6

The performance of the proposed model for 25-generations of UCI dataset (a) loss and (b) mean accuracy.

Figure 7

Accuracy for 50 epochs from UCI dataset.

Figure 8

Loss for 50 epochs from UCI dataset.

Figure 9

Confusion matrix for UCI dataset of proposed model.

Figure 10

UCI_HAR 10-Fold False Discovery Rate.

Figure 11

UCI_HAR Specificity.

Figure 12

Confusion matrix for Opportunity dataset of proposed model.

Figure 13

Accuracy of AUTO-CNN on Daphnet dataset.

Figure 14

Confusion matrix for Daphnet dataset of proposed model.

Figure 15

Daphnet 10-Fold False Discovery Rate.

Figure 16

Daphnet Specificity.

Figure 17

The best architecture for UCI dataset.

Figure 18

The best architecture for Daphnet dataset.