VFDS: Variational Foresight Dynamic Selection in Bayesian Neural Networks for Efficient Human Activity Recognition

03/31/2022
by   Randy Ardywibowo, et al.
0

In many machine learning tasks, input features with varying degrees of predictive capability are acquired at varying costs. In order to optimize the performance-cost trade-off, one would select features to observe a priori. However, given the changing context with previous observations, the subset of predictive features to select may change dynamically. Therefore, we face the challenging new problem of foresight dynamic selection (FDS): finding a dynamic and light-weight policy to decide which features to observe next, before actually observing them, for overall performance-cost trade-offs. To tackle FDS, this paper proposes a Bayesian learning framework of Variational Foresight Dynamic Selection (VFDS). VFDS learns a policy that selects the next feature subset to observe, by optimizing a variational Bayesian objective that characterizes the trade-off between model performance and feature cost. At its core is an implicit variational distribution on binary gates that are dependent on previous observations, which will select the next subset of features to observe. We apply VFDS on the Human Activity Recognition (HAR) task where the performance-cost trade-off is critical in its practice. Extensive results demonstrate that VFDS selects different features under changing contexts, notably saving sensory costs while maintaining or improving the HAR accuracy. Moreover, the features that VFDS dynamically select are shown to be interpretable and associated with the different activity types. We will release the code.

READ FULL TEXT

page 7

page 8

research
02/22/2017

Boosted Multiple Kernel Learning for First-Person Activity Recognition

Activity recognition from first-person (ego-centric) videos has recently...
research
01/01/2021

MoSen: Activity Modelling in Multiple-Occupancy Smart Homes

Smart home solutions increasingly rely on a variety of sensors for behav...
research
11/08/2018

BAR: Bayesian Activity Recognition using variational inference

Uncertainty estimation in deep neural networks is essential for designin...
research
07/31/2018

DFTerNet: Towards 2-bit Dynamic Fusion Networks for Accurate Human Activity Recognition

Deep Convolutional Neural Networks (DCNNs) are currently popular in huma...
research
04/01/2016

Leaving Some Stones Unturned: Dynamic Feature Prioritization for Activity Detection in Streaming Video

Current approaches for activity recognition often ignore constraints on ...
research
05/17/2018

Interpretable Parallel Recurrent Neural Networks with Convolutional Attentions for Multi-Modality Activity Modeling

Multimodal features play a key role in wearable sensor-based human activ...
research
08/12/2020

Implications on Feature Detection when using the Benefit-Cost Ratio

In many practical machine learning applications, there are two objective...

Please sign up or login with your details

Forgot password? Click here to reset