Underwater Object Classification and Detection: first results and open challenges

TL;DR

This paper reviews the challenges of underwater object detection, evaluates existing algorithms' shortcomings, and provides guidelines for future research, emphasizing the need for specialized architectures over conventional SOTA models.

Contribution

It offers a comprehensive analysis of underwater detection challenges and assesses the effectiveness of pretraining and detector types, highlighting the necessity for ad-hoc architectures.

Findings

Pretraining with ImageNet is not beneficial for underwater detection.

Two-stage detectors do not always outperform single-stage detectors in this context.

Underwater detection models struggle with lower quality datasets, indicating generalization issues.

Abstract

This work reviews the problem of object detection in underwater environments. We analyse and quantify the shortcomings of conventional state-of-the-art (SOTA) algorithms in the computer vision community when applied to this challenging environment, as well as providing insights and general guidelines for future research efforts. First, we assessed if pretraining with the conventional ImageNet is beneficial when the object detector needs to be applied to environments that may be characterised by a different feature distribution. We then investigate whether two-stage detectors yields to better performance with respect to single-stage detectors, in terms of accuracy, intersection of union (IoU), floating operation per second (FLOPS), and inference time. Finally, we assessed the generalisation capability of each model to a lower quality dataset to simulate performance on a real scenario, in which harsher conditions ought to be expected. Our experimental results provide evidence that underwater object detection requires searching for "ad-hoc" architectures than merely training SOTA architectures on new data, and that pretraining is not beneficial.