On March 6, at the entrance of the China Railway Science Research Institute, employees wore masks and hard hats and lined up one after another to "swipe their faces". They did not take off their masks or hats, they only stopped briefly in front of the screen, employee information and temperature conditions appeared on the screen, and the staff also achieved safe and fast entry.
Focus on the eye and optimize the attention algorithm model
"Face recognition is generally divided into two steps, one is face detection, and the other is face matching." Zhao Yang said that the recognition process is that the face machine first finds the face from the video image and then passes the face The facial features such as eyes, nose, mouth, eyebrows, etc., are identified in the personnel database by a certain algorithm.
Face masks face new challenges
"In response to the new needs for face recognition, if a large-scale hardware change is made to an existing smart construction site platform, on the one hand, it will increase equipment and increase costs, on the other hand, it will take a long time to transform and it will be difficult to meet the resumption of production." Zhao said The R & D team proposed a solution by strengthening software algorithms and upgrading the system platform.Focus on the eye and optimize the attention algorithm model
"Face recognition is generally divided into two steps, one is face detection, and the other is face matching." Zhao Yang said that the recognition process is that the face machine first finds the face from the video image and then passes the face The facial features such as eyes, nose, mouth, eyebrows, etc., are identified in the personnel database by a certain algorithm.
评论
发表评论