Deep Recurrent Neural Network for Multi-target Filtering

TL;DR

This paper introduces a deep recurrent neural network approach with a novel data association algorithm for improved multi-target filtering, capable of handling occlusions and dynamic target states.

Contribution

It presents a new adaptive learning framework combining LSTM-based RNNs and a data association algorithm for multi-target filtering with occlusion handling.

Findings

Promising results in simulation scenarios

Effective handling of occluded targets

Improved target state updates

Abstract

This paper addresses the problem of fixed motion and measurement models for multi-target filtering using an adaptive learning framework. This is performed by defining target tuples with random finite set terminology and utilisation of recurrent neural networks with a long short-term memory architecture. A novel data association algorithm compatible with the predicted tracklet tuples is proposed, enabling the update of occluded targets, in addition to assigning birth, survival and death of targets. The algorithm is evaluated over a commonly used filtering simulation scenario, with highly promising results.