Search Results (4)

The Chu-style lacquerware (CSL) dataset is a digital resource specifically developed for the digital preservation and inheritance of Chu-style lacquerware, which constitutes an important component of global intangible handicraft heritage. The dataset systematically integrates on-site photographic images from the Hubei Provincial Museum and official digital resources from the same institution, comprising 582 high-resolution images of Chu-style lacquerware, 72 videos of artifacts, and 37 images of traditional Chinese patterns. It comprehensively demonstrates the artistic characteristics of Chu-style lacquerware and provides support for academic research and cultural dissemination. The construction process of the dataset includes data screening, image standardization, Photoshop-based editing and adjustment, image inpainting, and image annotation. Based on this dataset, this study employs the Low-Rank Adaptation (LoRA) technique to train three core models and five style models, and systematically verifies the usability of the CSL dataset from five aspects. Experimental results show that the CSL dataset not only improves the accuracy and detail restoration of Artificial Intelligence (AI)-generated images of Chu-style lacquerware, but also optimizes the generative effect of innovative patterns, thereby validating its application value. This study represents the first dedicated dataset developed for AI generative models of Chu-style lacquerware. It not only provides a new technological pathway for the digital preservation and inheritance of cultural heritage, but also supports interdisciplinary research in archeology, art history, and cultural communication, highlighting the importance of cross-disciplinary collaboration in safeguarding and transmitting Intangible Cultural Heritage (ICH). Full article

This paper introduces TU-DAT, a novel, freely downloadable computer vision dataset for analyzing traffic accidents using roadside cameras. TU-DAT addresses the lack of public datasets for training and evaluating models focused on automatic detection and prediction of road anomalies. It comprises approximately 280 real-world and simulated videos, collected from traffic CCTV footage, news reports, and high-fidelity simulations generated using BeamNG.drive. This hybrid composition captures aggressive driving behaviors—such as tailgating, weaving, and speeding—under diverse environmental conditions. It includes spatiotemporal annotations and structured metadata such as vehicle trajectories, collision types, and road conditions. These features enable robust model training for anomaly detection, spatial reasoning, and vision–language model (VLM) enhancement. TU-DAT has already been utilized in experiments demonstrating improved performance of hybrid deep learning- and logic-based reasoning frameworks, validating its practical utility for real-time traffic monitoring, autonomous vehicle safety, and driver behavior analysis. The dataset serves as a valuable resource for researchers, engineers, and policymakers aiming to develop intelligent transportation systems that proactively reduce road accidents. Full article

Waste classification is a critical step in waste management that is time-consuming and necessitates automation to replace traditional approaches. Recently, machine learning (ML) and deep learning (DL) have gained attention from researchers seeking to automate waste classification by providing alternative computational techniques to address various waste-related challenges. Significant research on waste classification has emerged in recent years, reflecting the growing focus on this domain. This systematic literature review (SLR) explores the role of artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), in automating waste classification. Using Kitchenham’s and PRISMA guidelines, we analyze over 97 studies, categorizing AI-based techniques into ML-based, DL-based, and hybrid models. We further present an in-depth review of over fifteen publicly available waste classification datasets, highlighting key limitations such as dataset imbalance, real-world variability, and standardization issues. Our analysis reveals that deep learning and hybrid approaches dominate the current research landscape, with CNN-based architecture and transfer learning techniques showing particularly promising results. To guide future advancements, this study also proposes a structured roadmap that organizes challenges and opportunities into short-, mid-, and long-term priorities. The roadmap integrates insights on model accuracy, system efficiency, and sustainability goals to support the practical deployment of AI-powered waste classification systems. This work provides researchers with a comprehensive understanding of the state-of-the-art in ML and DL for waste classification and offers insights into areas that remain unexplored. Full article

Solar insecticidal lamps (SILs) are commonly used agricultural pest control devices that attract pests through a lure lamp and eliminate them using a high-voltage metal mesh. When integrated with Internet of Things (IoT) technology, SIL systems can collect various types of data, e.g., pest kill counts, meteorological conditions, soil moisture levels, and equipment status. However, the proper functioning of SIL-IoT is a prerequisite for enabling these capabilities. Therefore, this paper introduces the component composition and fault analysis of SIL-IoT. By examining long-term operational data from seven nodes deployed in real-world scenarios, different fault modes are identified. Six typical machine methods are adopted to verify the validity of the proposed dataset. The results indicate that machine learning algorithms can achieve high accuracy on the proposed dataset. Notably, voltage, current, and meteorological data play a crucial role in the fault diagnosis process for both SIL-IoT and other related agricultural IoT devices. Full article