Quick and energy-efficient Bayesian computing of binocular disparity using stochastic digital signals

TL;DR

This paper presents a novel probabilistic hardware approach using stochastic bitstreams to efficiently compute Bayesian binocular disparity, significantly reducing energy consumption and computational complexity in 3D scene reconstruction.

Contribution

It introduces a stochastic data representation method for Bayesian inference, enabling high-performance, energy-efficient disparity computation suitable for hardware implementation.

Findings

High performance in simulated stochastic implementation

Potential for energy-efficient hardware architectures

Applicability to sensorimotor processing and robotics

Abstract

Reconstruction of the tridimensional geometry of a visual scene using the binocular disparity information is an important issue in computer vision and mobile robotics, which can be formulated as a Bayesian inference problem. However, computation of the full disparity distribution with an advanced Bayesian model is usually an intractable problem, and proves computationally challenging even with a simple model. In this paper, we show how probabilistic hardware using distributed memory and alternate representation of data as stochastic bitstreams can solve that problem with high performance and energy efficiency. We put forward a way to express discrete probability distributions using stochastic data representations and perform Bayesian fusion using those representations, and show how that approach can be applied to diparity computation. We evaluate the system using a simulated stochastic implementation and discuss possible hardware implementations of such architectures and their potential for sensorimotor processing and robotics.