Memory-efficient particle filter recurrent neural network for object localization

TL;DR

This paper introduces a memory-efficient particle filter RNN architecture that improves object localization accuracy in noisy environments while using fewer parameters, making it suitable for environments of varying sizes.

Contribution

It combines particle filter ideas with GRU RNNs to create a memory-efficient model that outperforms previous approaches in noisy, challenging environments.

Findings

Provides more precise localization than competitors

Uses fewer trained parameters

Effective in symmetric and noisy environments

Abstract

This study proposes a novel memory-efficient recurrent neural network (RNN) architecture specified to solve the object localization problem. This problem is to recover the object states along with its movement in a noisy environment. We take the idea of the classical particle filter and combine it with GRU RNN architecture. The key feature of the resulting memory-efficient particle filter RNN model (mePFRNN) is that it requires the same number of parameters to process environments of different sizes. Thus, the proposed mePFRNN architecture consumes less memory to store parameters compared to the previously proposed PFRNN model. To demonstrate the performance of our model, we test it on symmetric and noisy environments that are incredibly challenging for filtering algorithms. In our experiments, the mePFRNN model provides more precise localization than the considered competitors and requires fewer trained parameters.