Latest Articles for Datasets

A new approach to understanding metaphors in videos through automated captioning.

A recent study replicates key findings on data interpretation using sound and visuals.

A study introduces a new benchmark for prompt performance in creating and retrieving images.

The ULS23 Challenge aims to improve tumor segmentation in CT scans for better cancer care.

A study on the effectiveness of various lightweight models in image classification.

A new dataset aids in spotting subjective content in Arabic news articles.

This study assesses GPT-4's ability to extract data from materials science literature.

A framework designed to standardize benchmarking in topological deep learning research.

A novel approach enhances detection of software security vulnerabilities using advanced models.

MedExQA sets a new standard for evaluating medical language models with a focus on explanations.

A new approach to predict mobile app UI changes based on user actions.

A new method enhances LLMs for generating high-quality UI code.

This study introduces a method to analyze complex biological datasets effectively.

OphNet enhances surgical workflow analysis with a rich video dataset.

Analyzing harmful memes and their effects on society.

Study examines the robustness of segmentation models against adversarial attacks in healthcare.

Study reveals bias differences in language models across various languages.

A new method improves detection of small moving targets in infrared images.

mOSCAR provides a multilingual dataset for improved AI understanding of text and images.

A comprehensive dataset merging images and text to aid machine learning.

New methods improve image datasets while ensuring privacy and performance.

A new benchmark tests compositional reasoning in advanced models.

This study examines audio methods for tracking pedestrian movement in urban areas.

Competition reveals vulnerabilities and defenses in language model security.

A new dataset improves the creation of foley audio for multimedia content.

A comprehensive dataset for Arabic handwritten text recognition and research.

New dataset enhances robots' grasping skills using natural language commands.

MMScan enhances AI’s ability to comprehend complex 3D environments with extensive annotations.

Researchers aim to improve machine understanding of daily activities through video analysis.

New study uses X-ray microtomography for detailed mouse brain analysis.

Research aims to clarify how we perceive different smells.

New dataset helps assess AI text accuracy and reliability.

New methods enhance clarity in healthcare question answering platforms.

A new dataset improves coherence in image-text sequences for effective content creation.

Examining the cultural nuances in interpreting Chinese Pun Rebus art.

A new method combines rule mining and link prediction to enrich knowledge graphs.

SEACrowd aims to improve AI representation for Southeast Asian languages and cultures.

This article discusses a new benchmark for combining images and text to find events in videos.

A method to enhance the identification of fake news using social media interactions.

IGL-Bench provides essential tools for better analyzing imbalanced graphs.