Artificial Intelligence

The dataset has undergone format conversion based on URPC2021_Sonar_images_data, enabling it to be trained by YOLO and RT-DETR models.

 The folder 'images' contains image files

The folder 'labels' contains TXT format annotation files.

The annotation file in the folder annotations is in XML format

Data.yaml is the configuration file for YOLO training

Data_deTR is the configuration file for RT-DETR and US-DETR training

This study presents a English-Luganda parallel corpus comprising over 2,000 sentence pairs, focused on financial decision-making and products. The dataset draws from diverse sources, including social media platforms (TikTok comments and Twitter posts from authoritative accounts like Bank of Uganda and Capital Markets Uganda), as well as fintech blogs (Chipper Cash and Xeno). The corpus covers a range of financial topics, including bonds, loans, and unit trust funds, providing a comprehensive resource for financial language processing in both English and Luganda.

In order to realize intelligent and accurate campus risk detection, this paper proposes an improved YOLOv10 algorithm that integrates Self-Calibrated Illumination algorithm. The algo-rithm optimizes the loss function by introducing an auxiliary bounding box, and accelerates model convergence. StarNet is employed to enhance the original network structure, feature extraction capability, and decrease parameter count and calculations.

Recent advances in generative visual content have led to a quantum leap in the quality of artificially generated Deepfake content. Especially, diffusion models are causing growing concerns among communities due to their ever-increasing realism. However, quantifying the realism of generated content is still challenging. Existing evaluation metrics, such as Inception Score and Fréchet inception distance, fall short on benchmarking diffusion models due to the versatility of the generated images.