Novel tracking approach based on fully-unsupervised disentanglement of the geometrical factors of variation

TL;DR

This paper introduces a fully unsupervised deep learning method that disentangles geometrical factors of variation for particle tracking, reducing the need for labeled data and manual tuning.

Contribution

It presents a novel autoencoder-based model that leverages geometric invariances for disentanglement, applicable to real detector data without supervision.

Findings

Effective disentanglement of geometric factors demonstrated on synthetic data

Model performs well on real particle tracking detector data

Requires multiple space transformations for meaningful disentanglement

Abstract

Efficient tracking algorithms are a crucial part of particle tracking detectors. While a lot of work has been done in designing a plethora of algorithms, these usually require tedious tuning for each use case. (Weakly) supervised Machine Learning-based approaches can leverage the actual raw data for maximal performance. Yet in realistic scenarios, sufficient high-quality labeled data is not available. While training might be performed on simulated data, the reproduction of realistic signal and noise in the detector requires substantial effort, compromising this approach. Here we propose a novel, fully unsupervised, approach to track reconstruction. The introduced model for learning to disentangle the factors of variation in a geometrically meaningful way employs geometrical space invariances. We train it through constraints on the equivariance between the image space and the latent representation in a Deep Convolutional Autoencoder. Using experimental results on synthetic data we show that a combination of different space transformations is required for meaningful disentanglement of factors of variation. We also demonstrate the performance of our model on real data from tracking detectors.