Potr-Aid :An AI-Driven Portrait Assistant for Professional-Quality Image Composition
Document Type
Conference Proceeding
Publication Date
4-2025
Abstract
In the contemporary landscape of ubiquitous portrait photography, achieving professional-quality results remains a significant challenge for many users. This poster introduces Portr-Aid, an AI-driven portrait assistant system designed to empower users to create high-quality portraits. Our system employs a modified ResNet50 architecture to analyze portrait shots, adhering to established compositional principles. The proposed method was trained and evaluated on a substantial dataset comprising over 60,000 portraits, demonstrating a precision of 0.9663 in identifying suboptimal portrait captures. Notably, the system exhibits high computational efficiency, processing images within 0.15 seconds on standard hardware. This research not only contributes to advancements in automated image analysis but also democratizes access to professional portrait techniques for users of all skill levels, with practical applications across both amateur and professional photography workflows. The implications of this work extend beyond personal photography, offering potential applications in educational settings, professional training programs, and automated photo editing software. Furthermore, the technology could significantly impact industries relying on high-quality portraits, from social media platforms and personal branding to e-commerce and digital marketing, potentially reducing the barriers between amateur and professional-grade photography.
Recommended Citation
Kahlon, Jaspal Singh, "Potr-Aid :An AI-Driven Portrait Assistant for Professional-Quality Image Composition" (2025). Student Research Symposium 2025. 37.
https://digitalcommons.tamusa.edu/srs_2025/37
Comments
Poster Session 1
3:30-5:00 p.m.
BLH Lobby