SmartParkAI: A Deep Learning and Computer Vision Framework for Parking Optimization in Metropolitan Environments

ICLR 2026 Conference Submission20596 Authors

Keywords: Deep Learning for Resource Allocation, Computer Vision (OpenCV), Transformer Architectures, Smart Mobility Systems, Parking Optimization

TL;DR: SmartParkAI is an AI-driven framework using deep learning, OpenCV, and transformers to optimize parking allocation, making it more adaptive and efficient than classical heuristic-based systems.

Abstract: Parking inefficiency is a major contributor to traffic congestion in metropolitan regions, with up to 40% of congestion caused by drivers searching for spaces. Classical parking management systems rely on static heuristics and lack adaptability, limiting their effectiveness. In contrast, our model integrates artificial intelligence, making it more efficient, scalable, and responsive to real-world conditions, when compared to classical models. This work presents SmartParkAI, a framework under development that leverages Deep Learning for Resource Allocation and Computer Vision (OpenCV) to model parking demand and supply in a two-sided marketplace. SmartParkAI employs Transformer Architectures for dynamic allocation and time optimization, while visual models perform object detection and segmentation for compliance verification. By processing heterogeneous signals—including geospatial data, server metadata, and community violation reports, the system continuously adapts to evolving conditions. SmartParkAI thus advances Smart Mobility Systems and offers a practical pathway for large-scale Parking Optimization in densely populated environments.

Primary Area: generative models

Submission Number: 20596