A Safety Assurable Human-Inspired Perception Architecture

TL;DR

This paper proposes a human-inspired dual process perception architecture to enhance safety, interpretability, and robustness of AI perception systems, addressing key limitations of deep neural networks in autonomous applications.

Contribution

It introduces a novel dual process architecture inspired by human cognition to improve safety and assurance in AI perception systems, combining fast and slow processing modes.

Findings

Sketches a perception architecture based on human cognition models

Highlights how existing work addresses parts of the architecture

Identifies future research directions for safety assurance

Abstract

Although artificial intelligence-based perception (AIP) using deep neural networks (DNN) has achieved near human level performance, its well-known limitations are obstacles to the safety assurance needed in autonomous applications. These include vulnerability to adversarial inputs, inability to handle novel inputs and non-interpretability. While research in addressing these limitations is active, in this paper, we argue that a fundamentally different approach is needed to address them. Inspired by dual process models of human cognition, where Type 1 thinking is fast and non-conscious while Type 2 thinking is slow and based on conscious reasoning, we propose a dual process architecture for safe AIP. We review research on how humans address the simplest non-trivial perception problem, image classification, and sketch a corresponding AIP architecture for this task. We argue that this architecture can provide a systematic way of addressing the limitations of AIP using DNNs and an approach to assurance of human-level performance and beyond. We conclude by discussing what components of the architecture may already be addressed by existing work and what remains future work.