Frontal Face GAN : Generating Pose-Corrected Human Faces

Ayush R. Barai; Santosh G. Giram; Pratik N. Salve; Harshad H. Pawar; Omprakash K.Parihar

Frontal Face GAN : Generating Pose-Corrected Human Faces

Author(s)	Ayush R. Barai, Santosh G. Giram, Pratik N. Salve, Harshad H. Pawar, Omprakash K.Parihar
Country	India
Abstract	Generating identity-preserving frontal face images from unconstrained real-world profile views represent a fascinating problem in computer vision, the solution of which is critical for enhancing downstream tasks such as face recognition and attribute analysis. Current approaches often struggle with balancing the realistic requirements of 3D geometrical fidelity against real-world-nuisance variations such as uncontrolled light, occlusions, and heterogeneous backgrounds. This paper presents a novel two-stage transfer learning approach for training a Generative Adversarial Network (GAN) for the task. Our model, which follows the Pix2pix architecture with a U-Net generator and a PatchGAN discriminator, is first pre-trained on the CMU Multi-PIE dataset, which is controlled and makes the model learn the necessary 3D-aware geometric transformations from facial rotation. Thereafter, the pre-trained model is fine-tuned on realistic difficulties presented by the CFPW dataset. In the interest of practical application, our system integrates an MTCNN face detector for the automated pre-processing of ’wild’ images. According to our study, this two-phase training setting turns out to be very powerful, yielding a robust model capable of generating high-fidelity geometrically correct frontal images from unconstrained challengingly posed profile images.
Keywords	{(Face Frontalization, Encoder, Decoder, Generator, Discriminator, Generative Adversarial Networks (GANs))}
Published In	Volume 7, Issue 2, March-April 2026
Published On	2026-04-21

View / Download PDF File

E-ISSN 3048-7641

doi

CrossRef DOI is assigned to each research paper published in our journal.

AIJFR DOI prefix is
10.63363/aijfr

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 7 Isu 2 Cover Page Vol 7 Isu 1 Cover Page Vol 6 Isu 6

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About AIJFR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-79-39-59	editor@aijfr.com

Advanced International Journal for Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Frontal Face GAN : Generating Pose-Corrected Human Faces

Share this