Abstract:Current sensing methods often ignore the fact that their sensing targets are dynamic and can change over time. As a result, to build an accurate model should not always be of the first priority. What we need is to establish an adaptive modelling framework. Lack of this adaptability hinders us from building a more intelligent sensing system. In this paper, we try to apply inspirations from human cognition to design a more intelligent sensing and modelling system, which can adaptively detect anomalies. Based on our understanding of free-energy and Infomax principle, the target of sensing and modelling is not to get as much data as possible, or to build the most accurate model, but to establish an adaptive representation of target and achieve balance between sensing performance and system resource consumption. Formally speaking, from the perspective of free energy minimization, this corresponds to a balance between accuracy and the minimization of complexity costs. To achieve this goal, we adopt a working memory mechanism to help the model evolve with the target; we use deep autoencoder network as model representation, which models complex data with its nonlinear and hierarchical architecture. Since we typically only have partial observations from sensed target, we design a variance of autoencoder that can reconstruct corrupted input. We utilize attentional surprise mechanism to control model update. Training of the deep network is driven by surprises detected (anomalies), which indicates model failure or target's new behaviour. Due to partial observations, we are not able to minimize free-energy in a single update, but iteratively minimize it by finding new optimization bounds. While both random and non-random sensor selection can create new optimization bounds, non-random methods like surprise minimization used in this paper demonstrate better performance. In our system, the model update frequency is controlled by several parameters, including surprise threshold and memory size. These parameters control the alertness as well as the resource consumption of the system in a top-down manner. For evaluation, we conducted experiments on simulated data to test whether our methodology makes the model more adaptive. The result showed that we achieved this aim. We also applied our method to a real application, which is EEG (Electroencephalography) seizure detection. This application shows features that we desired.

An Adaptive Multi-Sensor Visual Attention Model

Cognitive Sensing: Adaptive Anomalies Detection with Deep Networks

Visual Attention Modeling Based on Short-Term Environmental Adaption

An Attention Selection Model With Visual Memory And Online Learning

Adaptive Attention Module for Image Recognition Systems in Autonomous Driving

MAM: A Multipath Attention Mechanism for Image Recognition

A Multi-Sensor Mapping Bi-LSTM Model of Bridge Monitoring Data Based on Spatial-Temporal Attention Mechanism

An Object-Based Visual Attention Model for Robotic Applications

A Biologically Inspired Object-Based Visual Attention Model

A Neural Network Model of Visual Attention Integrating Biased Competition and Reinforcement Learning

Classification Model of Visual Attention Based on Eye Movement Data

Learn to Pay Attention Via Switchable Attention for Image Recognition

ATCON: Attention Consistency for Vision Models

Object detection based on an adaptive attention mechanism

Multimodal Continuous Visual Attention Mechanisms

General Recurrent Attention Model for Jointly Multiple Object Recognition and Weakly Supervised Localization.

Switchable Self-attention Module

A Multiscale Self-Adaptive Attention Network for Remote Sensing Scene Classification

Visual Attention Model with a Novel Learning Strategy and Its Application to Target Detection from SAR Images.

An Autonomous Visual Perception Model For Robots Using Object-Based Attention Mechanism

A Dorsal Pathway Guided Visual Attention Model