Advancing Image Understanding in Poor Visibility Environments: A Collective Benchmark Study

By Wenhan Yang, Ye Yuan, Wenqi Ren, Jiaying Liu, Walter J. Scheirer, Zhangyang Wang, Taiheng Zhang, Qiaoyong Zhong, Di Xie, Shiliang Pu and 57 more. IEEE Transactions on Image Processing, 2019

337 citations recorded by the source corpus.

Existing enhancement methods are empirically expected to help the high-level end computer vision task: however, that is observed to not always be the case in practice. We focus on object or face detection in poor visibility enhancements caused by bad weathers (haze, rain) and low light conditions. 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 annotated objects/faces. We launched the UG2+ challenge Track 2 competition in IEEE CVPR 2019, aiming to evoke a comprehensive discussion and exploration about whether and how low-level vision techniques can benefit the high-level automatic visual recognition in various scenarios. 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. Thanks to a large participation from the research community, we are able to analyze representative team solutions, striving to better identify the strengths and limitations of existing mindsets as well as the future directions.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/9ac5bb2be62309abc73f5e78839e807101db7ecc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement (2020, methodology): “3 demonstrated three groups of visual comparisons on DarkFace dataset [41].”
Ultra-High-Definition Low-Light Image Enhancement: A Benchmark and Transformer-Based Method (2022, methodology): “To verify whether LLIE is beneﬁcial for downstream tasks, we randomly select 300 images from the DARK FACE dataset (Yang et al. 2020) and pre-process these images using the top three methods in our benchmark study.”
EnlightenGAN: Deep Light Enhancement Without Paired Supervision (2019, background): “Image enhancement as pre-processing for improving subsequent high-level vision tasks has recently received increasing attention [41], [53]–[55], with a number of benchmarking efforts [52], [56]–[58].”
Toward Fast, Flexible, and Robust Low-Light Image Enhancement (2022, positive): “Here, we tested lots of challenging in-the-wild examples from the DARK FACE [31] and ExDark [15] datasets.”
PSD: Principled Synthetic-to-Real Dehazing Guided by Physical Priors (2021, background): “As pointed out by several recent studies [19,33,40], the performance of high-level computer vision tasks, such as object detection and recognition, will severely deteriorate in the presence of environmental degradations.”
Learning to Enhance Low-Light Image via Zero-Reference Deep Curve Estimation (2021, methodology): “9) using the evaluation tool(3) provided in DARK FACE dataset [42].”