First Step toward Model-Free, Anonymous Object Tracking with Recurrent Neural Networks

TL;DR

This paper introduces a novel, end-to-end trainable neural network approach for anonymous object tracking that performs well across various noisy and dynamic scenarios, differing from traditional filtering and detection methods.

Contribution

It presents a new model combining convolutional and recurrent networks trained offline to track anonymous objects in noisy environments, addressing limitations of existing methods.

Findings

Effective in diverse scenarios with varying object counts and noise levels

Robust to mismatches between training and test object shapes

Outperforms traditional filtering and detection-based tracking methods

Abstract

In this paper, we propose and study a novel visual object tracking approach based on convolutional networks and recurrent networks. The proposed approach is distinct from the existing approaches to visual object tracking, such as filtering-based ones and tracking-by-detection ones, in the sense that the tracking system is explicitly trained off-line to track anonymous objects in a noisy environment. The proposed visual tracking model is end-to-end trainable, minimizing any adversarial effect from mismatches in object representation and between the true underlying dynamics and learning dynamics. We empirically show that the proposed tracking approach works well in various scenarios by generating artificial video sequences with varying conditions; the number of objects, amount of noise and the match between the training shapes and test shapes.