BrackishMOT: The Brackish Multi-Object Tracking Dataset

TL;DR

BrackishMOT introduces a new underwater multi-object tracking dataset focused on small fish in turbid environments, along with baseline results and a synthetic data generation framework to improve tracking performance.

Contribution

The paper presents the first annotated underwater MOT dataset for turbid environments, a synthetic data creation framework, and baseline tracking results.

Findings

Synthetic data improves tracking accuracy.

Combining real and synthetic data enhances performance.

BrackishMOT fills a gap in underwater MOT datasets.

Abstract

There exist no publicly available annotated underwater multi-object tracking (MOT) datasets captured in turbid environments. To remedy this we propose the BrackishMOT dataset with focus on tracking schools of small fish, which is a notoriously difficult MOT task. BrackishMOT consists of 98 sequences captured in the wild. Alongside the novel dataset, we present baseline results by training a state-of-the-art tracker. Additionally, we propose a framework for creating synthetic sequences in order to expand the dataset. The framework consists of animated fish models and realistic underwater environments. We analyse the effects of including synthetic data during training and show that a combination of real and synthetic underwater training data can enhance tracking performance. Links to code and data can be found at https://www.vap.aau.dk/brackishmot