Synthesizing Reality: Leveraging the Generative AI-Powered Platform Midjourney for Construction Worker Detection

2025-07-17

AI Summary

In construction-site personnel detection, the scarcity of accurately annotated real-world data severely limits deep learning performance. To address this, this paper proposes a synthetic data generation method leveraging Midjourney—a text-to-image generative AI—via systematic prompt engineering to produce 12,000 high-diversity, high-fidelity images with pixel-level annotations. The synthetic dataset is rigorously validated through human verification and downstream deep neural network (DNN) training for object detection. Experimental results demonstrate substantial performance gains: on a real-world construction test set, the model achieves AP@0.5 = 0.937 and AP@0.5:0.95 = 0.642; on the synthetic test set, it attains AP@0.5 = 0.994 and AP@0.5:0.95 = 0.919. This work represents the first large-scale application of Midjourney in industrial vision data synthesis, empirically establishing generative AI’s efficacy—and practical limitations—in mitigating real-data scarcity under low-shot conditions. It delivers a reproducible, scalable data paradigm for AI deployment in construction automation.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Web Mining and Content Analysis: Web data generation and simulation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

While recent advancements in deep neural networks (DNNs) have substantially enhanced visual AI's capabilities, the challenge of inadequate data diversity and volume remains, particularly in construction domain. This study presents a novel image synthesis methodology tailored for construction worker detection, leveraging the generative-AI platform Midjourney. The approach entails generating a collection of 12,000 synthetic images by formulating 3000 different prompts, with an emphasis on image realism and diversity. These images, after manual labeling, serve as a dataset for DNN training. Evaluation on a real construction image dataset yielded promising results, with the model attaining average precisions (APs) of 0.937 and 0.642 at intersection-over-union (IoU) thresholds of 0.5 and 0.5 to 0.95, respectively. Notably, the model demonstrated near-perfect performance on the synthetic dataset, achieving APs of 0.994 and 0.919 at the two mentioned thresholds. These findings reveal both the potential and weakness of generative AI in addressing DNN training data scarcity.

Problem

• Addressing inadequate data diversity in construction worker detection
• Generating synthetic images using Midjourney for DNN training
• Evaluating model performance on real and synthetic construction datasets

Innovation

• Generative AI platform Midjourney for synthetic images
• 3000 prompts create 12000 diverse realistic images
• Synthetic dataset trains DNN for worker detection