Tag: multimodal AI
Browse our exclusive articles!
K-MetBench: Benchmarking AI for Korean Meteorology
Discover K-MetBench, a multidimensional benchmark for evaluating AI expert reasoning, locality, and multimodality in Korean meteorology applications.
Enhancing VLM Reasoning with Visual Cues & Reflection
Discover how ForeSight boosts Vision-Language Models' reasoning using low-level visual cues and mask-based feedback for superior AI understanding.
Prompted Weak Supervision for Meme Hate Speech Detection
Discover how prompted weak supervision improves detection of homophobic and transphobic hate speech in memes across multiple languages.
Audio Hallucination Challenges in Egocentric Video AI
Discover how audio hallucinations affect egocentric video AI models and the need for better evaluation to improve accuracy in multimodal understanding.
Vision-Language-Action in Robotics: Key Datasets & Benchmarks
Explore the latest survey on Vision-Language-Action in robotics, covering datasets, benchmarks, and data engines driving embodied learning advancements.