Face detection in low light scenarios is challenging but vital to manypractical applications, e.g., surveillance video, autonomous driving at night . The gap between normal and lowlight is too huge and complex for both pixel-level and object-level . We propose a joint High-LowAdaptation (HLA) framework to address the issue . Through a bidirectional low-level adaptation andmulti-task high-level adaptations scheme, our HLA-Face outperformsstate-of-the-art methods even without using dark face labels for training . Our project is publicly available at https://daooshee.io/HLA-Face-Website/ .

Author(s) : Wenjing Wang, Wenhan Yang, Jiaying Liu

Links : PDF - Abstract

Code :

Keywords : face - level - hla - high - detection -

Leave a Reply

Your email address will not be published. Required fields are marked *