Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{201722,
        author = {B Aysha Banu and Samundeeswari K and CS Ahamed Maajidh and S Mohaideen Ajmeer Dhulkarunai and S Mohamed Mafthoon},
        title = {A HYBRID FACE DETECTION SYSTEM OPTIMIZED VIA LATENT DIFFUSION SYNTHETIC IMAGE SYNTHESIS},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {5196-5200},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=201722},
        abstract = {The proliferation of surveillance systems, identity verification platforms, and augmented-reality applications demands face detectors that remain robust under extreme pose variation, occlusion, low illumination, and severe class imbalance. This paper presents a Hybrid Face Detection System (HFDS) that synergises five specialised deep-learning models—RetinaFace++, DiffYOLO-Face, DETR-SynFace, GAN-Guard, and FusionNet-X—within a unified ensemble framework. To overcome the chronic scarcity of annotated edge-case imagery, we introduce Latent Diffusion Synthetic Image Synthesis (LDSS), a novel data-augmentation pipeline that leverages a fine-tuned latent diffusion model to generate photorealistic, semantically consistent face images spanning underrepresented demographics, lighting conditions, and partial-occlusion patterns. Experimental evaluation on WIDERFACE, FDDB, and our LDSS-augmented benchmark demonstrates a mean Average Precision (mAP) of 96.8%, a 3.1-point improvement over the previous state-of-the-art single-model detectors. The system sustains 28 FPS on consumer-grade hardware, confirming real-time viability.},
        keywords = {Hybrid Face Detection, Latent Diffusion Models, Synthetic Image Synthesis, Retina Face, DETR, GAN, Ensemble Learning, Fusion Net-X, WIDER FACE, mAP},
        month = {May},
        }