YOLO and Faster R-CNN Object Detection in Architecture, Engineering and Construction (AEC): Applications, Challenges, and Future Prospects

SSRN Electronic Journal, Год журнала: 2023, Номер unknown

Опубликована: Янв. 1, 2023

Object detection plays a crucial role in transforming the Architecture, Engineering, and Construction (AEC) industry, enhancing project efficiency, safety, overall productivity. This study explores applications, challenges, future potential of two cutting-edge object algorithms, namely You Only Look Once (YOLO) Faster Region-based Convolutional Neural Networks (Faster R-CNN), within realm AEC. The research comprehensively investigates diverse applications YOLO R-CNN AEC, including real-time site monitoring, structural integrity assessment, safety protocol enforcement, automated progress tracking, quality control. These algorithms have propelled AEC industry forward, enabling advancements autonomous inspection, defect detection, resource management. Consequently, these innovations enhanced decision-making processes optimized lifecycles. Nevertheless, integrating technologies presents challenges. paper meticulously examines hurdles such as data annotation complexities, algorithmic limitations, computational demands. It also delves into ethical considerations, privacy, cybersecurity concerns, shedding light on implications associated with widespread adoption industry. Looking ahead, outlines prospects discusses solutions to existing include development more robust streamlined processes, edge computing. Moreover, emerging trends like Explainable AI (XAI) Generative Adversarial (GANs), envisioning their integration for even sophisticated provides valuable insights researchers, practitioners, policymakers, paving way efficient, innovative, ethically responsible sector.

Язык: Английский

---

Application of Large Language Models in Medical Training Evaluation: Can ChatGPT Be a Standardized Patient? An Exploratory Study (Preprint)

Опубликована: Апрель 11, 2024

With the increasing interest of Large Language Models’ (LLMs) application in medical field, feasibility its potential usage as a Standardized Patient (SP) assessment is rarely evaluated. Specifically, we delved into using ChatGPT, representative LLM, transforming education by serving cost-effective alternative to SPs, specifically for history-taking tasks. The study aims explore ChatGPT's viability and performance an SP, employing prompt engineering refine accuracy utility assessments. A two-phase experiment was designed assess SP education. first phase tested through simulating conversations on Inflammatory Bowel Disease (IBD), categorizing responses poor, medium, good inquiries based relevance accuracy. For second phase, more structured used detailed scripts evaluate against specific criteria, focusing anthropomorphism, clinical accuracy, adaptability. Adjustments were made prompts response shortcomings, with comparative analysis ChatGPT’s between original revised track improvements. methodology included statistical ensure rigorous evaluation, data collected November December 2023. test confirmed ability simulate effectively, differentiating varying degrees Score differences poor (74.7, SD=5.44) medium (82.67, SD=5.30) inquiry groups (P < .001), (85, SD=3.27) .001) significant at significance level α = .05, while score not statistically (P= .158). took 90 runs. However, ideal without proper restriction. Subsequent enhancements, including use prompts, instructed ChatGPT avoid jargon realism, provide accurate concise improve grading adaptability following prompts. total number trials experimental 300. significantly improved adaptability, leading marked reduction scoring discrepancies. 4.926 times compared unrevised prompt. difference percentage (SDP) drops from 29.83% 6.06%, drop standard deviation 0.55 0.068. viable tool SPs assessments, enhance training. By incorporating targeted realism improve, approaching actual use. despite promising outcomes, continuous refinement essential fully establish LLM’s (such ChatGPT) reliability settings.

Язык: Английский

---

Transition to Sustainable Human-Centric Education in Emerging Artificial Intelligence Industry 5.0

Advances in higher education and professional development book series, Год журнала: 2024, Номер unknown, С. 37 - 76

Опубликована: Дек. 18, 2024

The wave of European Commission conception for the Fifth Industrial Revolution, or Industry 5.0, a new paradigm towards sustainability that will use technology to transform world is essentially proliferating. In this paper, authors exceptionally advanced generative artificial intelligence (AI) in development conversational education pedagogy Society 5.0 by describing technological underpinnings, guiding principles, essential values, and key components education. Conversational AI was projected with help user-centric ChatGPT-5 .The goal provide educational technologist best practices experimentally validated guidelines measuring, enhancing, maintaining human centeredness. It anticipated would incorporate more sophisticated multimodal features, allowing it process produce text addition images, voice, maybe video. Creating intricate visual material, helping video content offering dynamic captivating user experiences

Язык: Английский

---

The role of large pre-trained models in ecology and biodiversity conservation: Opportunities and Challenges

Authorea (Authorea), Год журнала: 2024, Номер unknown

Опубликована: Март 25, 2024

Large pre-trained models (LPMs) have the capabilities to understand natural language, code, and diverse data including images; e.g., large language (LLMs), code-generative models, vision (LVMs) as well combined multi-modal models. We outlines main applications of LPMs for ecology biodiversity conservation. These include generating ecological data, providing insights into public opinion sentiment. highlighted significant potential use Ecology-specialized They offer unprecedented opportunities analyzing extracting meaningful insights, informing conservation decisions.

Язык: Английский

---

AI-Driven Carbon Emissions Tracking and Mitigation Model

2021 IST-Africa Conference (IST-Africa), Год журнала: 2024, Номер unknown, С. 1 - 8

Опубликована: Май 20, 2024

Язык: Английский

---