AI研究論文每日精選

VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasters

Mini-Omni: Language Models Can Hear, Talk While Thinking in Streaming