Dana

Abstract

Motion sensors embedded in wearable and mobile devices allow for dynamic selection of sensor streams and sampling rates, enabling several applications, such as power management and data-sharing control. While deep neural networks (DNNs) achieve competitive accuracy in sensor data classification, DNN architectures generally process incoming data from a fixed set of sensors with a fixed sampling rate, and changes in the dimensions of their inputs cause considerable accuracy loss, unnecessary computations, or failure in operation. To address these limitations, we introduce a dimension-adaptive pooling (DAP) layer that makes DNNs flexible and more robust to changes in sensor availability and in sampling rate. DAP operates on convolutional filter maps of variable dimensions and produces an input of fixed dimensions suitable for feedforward and recurrent layers. Further, we propose a dimension-adaptive training (DAT) procedure for enabling DNNs that use DAP to better generalize over the set of feasible data dimensions at inference time. DAT comprises the random selection of dimensions during the forward passes and optimization with accumulated gradients of several backward passes. Combining DAP and DAT, we show how to transform existing non-adaptive DNNs into a Dimension-Adaptive Neural Architecture (DANA), while keeping the same number of parameters. Compared to existing approaches, our solution provides better average classification accuracy over the range of possible data dimensions at inference time and does not require up-sampling or imputation, thus reducing unnecessary computations. Experimental results on seven datasets (four benchmark real-world datasets for human activity recognition and three synthetic datasets) show that DANA prevents significant losses in classification accuracy of the state-of-the-art DNNs and, compared to baselines, it better captures correlated patterns in sensor data under dynamic sensor availability and varying sampling rates.

A Trainable Open-Source Machine Learning Accelerometer Activity Recognition Toolbox: Deep Learning Approach

“… Course for accelerometer data collection on the campus of the Queen Mary University of London for the MotionSense data set; graph from Malekzadeh et al [ 26 ]. …”

“…Malekzadeh et al [ 26 ] proposed a new model, which tries to counteract the aforementioned shortcomings by introducing a dimension-adaptive pooling (DAP) layer, which makes DNNs robust to changes in not only sampling rates but also dimensional changes of the data due to varying sensor availability.…”

Domain Adaptation for Sensor-Based Human Activity Recognition with a Graph Convolutional Network

“…A l-layer GCN means that it is composed of l layers, while each convolutional layer constructs the embedding l of each node through the previous layer, as shown in (2). The embedding of N nodes at the l-th layer as X (l+1) ∈ R N×F l .…”

“…This approach involves the selection of diverse datasets from websites spanning various domains, followed by the utilization of transfer learning techniques to establish user behavior models by leveraging the data from these domains. On the other hand, [2] introduced the Dimension-Adaptive Neural Architecture (DANA), a novel solution that empowers deep neural networks to dynamically adapt to changes in input data dimensions. This adaptation capability allows DANA to effectively address challenges such as sensor availability and adaptive sampling during inference.…”