3D face recognition algorithm based on deep Laplacian pyramid under the normalization of epidemic control.

Area of Science:

• Computer Vision
• Biometrics
• Artificial Intelligence

Background:

• The normalization of epidemic control necessitates rapid and accurate face recognition for public safety.
• Existing methods may struggle with accuracy and speed in diverse environmental conditions.

Purpose of the Study:

• To propose an innovative Laplacian pyramid algorithm for deep 3D face recognition.
• To achieve fast and high-precision identification suitable for public use during epidemic control.

Main Methods:

• Utilized a 2D to 3D structure representation for correlating crucial point information and dense alignment.
• Constructed a five-layer Laplacian depth network based on a 3D critical point model.
• Employed multi-scale residual weights in the loss function and an end-to-end cascade design for real-time performance.

Main Results:

• Achieved high-precision 3D face recognition through multi-scale, multi-modal mapping and reconstruction.
• Demonstrated robustness in harsh, low-light, and noisy environments, recognizing various skin colors and postures.
• Ensured fast personnel screening while maintaining identification accuracy.

Conclusions:

• The proposed Laplacian pyramid algorithm offers a robust and efficient solution for 3D face recognition.
• This method supports public safety initiatives under normalized epidemic control conditions.
• The algorithm's adaptability makes it suitable for real-world applications with challenging environmental factors.