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.

Rethinking Saliency Map: An Context-aware Perturbation Method to Explain EEG-based Deep Learning Model

Rethinking Saliency Map: A Context-Aware Perturbation Method to Explain EEG-Based Deep Learning Model

Cognitive Sensing: Adaptive Anomalies Detection with Deep Networks

Explain EEG-based End-to-end Deep Learning Models in the Frequency Domain

Electroencephalogram Emotion Recognition Based on Empirical Mode Decomposition and Optimal Feature Selection.

What Do Deep Saliency Models Learn about Visual Attention?

Counterfactual-based Saliency Map: Towards Visual Contrastive Explanations for Neural Networks.

Identifying EEG Biomarkers of Depression with Novel Explainable Deep Learning Architectures

A Saliency based Feature Fusion Model for EEG Emotion Estimation

A Psychophysically Oriented Saliency Map Prediction Model

Identifying Reproducibly Important EEG Markers of Schizophrenia with an Explainable Multi-Model Deep Learning Approach

Evaluation of Interpretability for Deep Learning algorithms in EEG Emotion Recognition: A case study in Autism

Concept-based explainability for an EEG transformer model

Graphical Perception of Saliency-based Model Explanations

Evaluating Saliency Map Explanations for Convolutional Neural Networks: A User Study

Learning Reliable Visual Saliency for Model Explanations

Emotion-Agent: Unsupervised Deep Reinforcement Learning with Distribution-Prototype Reward for Continuous Emotional EEG Analysis

A Novel Method to Study Bottom-up Visual Saliency and its Neural Mechanism

Perturbing a Neural Network to Infer Effective Connectivity: Evidence from Synthetic EEG Data

An attention-based hybrid deep learning model for EEG emotion recognition

An Emotion Analysis Method That Integrates Eeg Signals And Music Therapy Using A Transformer Model