A survey on computer vision based human analysis in the COVID-19 era

Area of Science:

• Computer Vision
• Artificial Intelligence
• Public Health

Background:

• The COVID-19 pandemic introduced challenges for computer vision, particularly in analyzing individuals due to face masks.
• Prevention measures like masks impacted human analysis techniques such as face recognition and detection.

Purpose of the Study:

• To survey the impact of COVID-19 on computer vision-based human analysis.
• To review methods addressing challenges posed by facial masks in computer vision tasks.
• To discuss datasets, open challenges, and future directions in the field.

Main Methods:

• Comprehensive literature review of computer vision research related to COVID-19.
• Analysis of studies focusing on the effects of facial masks on human analysis algorithms.
• Review of relevant datasets for developing and evaluating computer vision models.

Main Results:

• Facial masks significantly affect the performance of various computer vision techniques for human analysis.
• Recent research has proposed several solutions to mitigate the performance degradation caused by masks.
• Existing datasets are reviewed for their utility in COVID-19 related computer vision applications.

Conclusions:

• Computer vision techniques for human analysis require adaptation to function effectively with facial occlusions.
• Further research is needed to address open challenges and advance the field, especially for real-world public health applications.
• This survey provides a valuable resource for researchers and the public interested in computer vision and its role during the pandemic.