Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Locally linear regression for pose-invariant face recognition.

Xiujuan Chai1, Shiguang Shan, Xilin Chen

  • 1Harbin Institute of Technology, Harbin 150001, China. xjchai@jdl.ac.cn

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|July 4, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Multi-View Analysis of Facial Expressions in Minimally Verbal Autism: Preliminary Evidence From Social Communication Observations.

Journal of autism and developmental disorders·2026
Same author

VLBiasBench: A Comprehensive Benchmark for Evaluating Bias in Large Vision-Language Model.

IEEE transactions on pattern analysis and machine intelligence·2026
Same author

Revisiting Face Forgery Detection: From Facial Representation to Forgery Detection.

IEEE transactions on pattern analysis and machine intelligence·2026
Same author

Distillation-SAM: Knowledge Distillation-Based Auto-Prompt Embedding Learning for Surgical Image Segmentation.

IEEE transactions on medical imaging·2026
Same author

Toward Transferable Defense Against Malicious Image Edits.

IEEE transactions on pattern analysis and machine intelligence·2026
Same author

Dynamic Attention Analysis for Backdoor Detection in Text-to-Image Diffusion Models.

IEEE transactions on pattern analysis and machine intelligence·2025

This study introduces a novel locally linear regression (LLR) method to generate virtual frontal face images from nonfrontal views. This approach significantly improves face recognition systems by overcoming viewpoint variations.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Biometrics

Background:

  • Viewpoint variation is a major challenge for face recognition systems.
  • Generating a virtual frontal view from nonfrontal images is a potential solution.

Purpose of the Study:

  • To propose a simple, efficient, and novel locally linear regression (LLR) method for virtual frontal face generation.
  • To address the bottleneck of viewpoint variation in face recognition.

Main Methods:

  • The method assumes an approximate linear mapping between nonfrontal and frontal face images.
  • Globally linear regression is used to estimate this mapping.
  • Locally Linear Regression (LLR) is proposed for improved accuracy with coarse alignment, using dense sampling of local patches and applying regression to predict virtual frontal patches.

Related Experiment Videos

Main Results:

  • The proposed LLR method effectively generates virtual frontal face views.
  • Experimental results on the CMU PIE database demonstrate a distinct advantage over the Eigen light-field method.

Conclusions:

  • The LLR method offers a robust solution for generating virtual frontal faces.
  • This technique can significantly enhance the performance of face recognition systems under varying poses.