UG$^{2+}$ Track 2: A Collective Benchmark Effort for Evaluating and Advancing Image Understanding in Poor Visibility Environments

TL;DR

This paper introduces a comprehensive benchmark for evaluating object and face detection in poor visibility conditions like haze, rain, and low light, highlighting the challenges and need for further technical improvements.

Contribution

It presents the first large-scale, real-world benchmark datasets for detection in adverse weather and lighting, facilitating fair comparison and encouraging research advancements.

Findings

Existing enhancement methods do not always improve detection performance.

Baseline results show high difficulty of the new datasets.

Large room for innovation in low-visibility image understanding.

Abstract

The UG$^{2+}$ challenge in IEEE CVPR 2019 aims to evoke a comprehensive discussion and exploration about how low-level vision techniques can benefit the high-level automatic visual recognition in various scenarios. In its second track, we focus on object or face detection in poor visibility enhancements caused by bad weathers (haze, rain) and low light conditions. While existing enhancement methods are empirically expected to help the high-level end task, that is observed to not always be the case in practice. To provide a more thorough examination and fair comparison, we introduce three benchmark sets collected in real-world hazy, rainy, and low-light conditions, respectively, with annotate objects/faces annotated. To our best knowledge, this is the first and currently largest effort of its kind. Baseline results by cascading existing enhancement and detection models are reported, indicating the highly challenging nature of our new data as well as the large room for further technical innovations. We expect a large participation from the broad research community to address these challenges together.