New AI Research Models by Meta to Foster Innovation
New AI Research Models by Meta to Foster Innovation
Meta's Fundamental AI Research team is unveiling several advanced AI models to propel innovation and global collaboration. Among these are models for image-to-text and text-to-music generation, multi-token prediction, and AI-generated speech detection, aimed at advancing AI responsibly.
Chameleon: Versatile Mixed-Modal AI
Chameleon is a groundbreaking model capable of processing and generating both text and images simultaneously. This unique capability allows for creative applications like generating captions or creating new scenes from mixed inputs.
Multi-Token Prediction for Efficient AI
To enhance the efficiency of language models, Meta introduces multi-token prediction. This innovative approach trains AI to predict multiple words simultaneously, improving speed and accuracy in text generation.
JASCO and AudioSeal Innovations
JASCO offers improved control over AI music generation by integrating various inputs like chords and beats, while AudioSeal introduces an advanced method for detecting AI-generated speech, optimizing detection speed and efficiency.
Promoting Diversity in AI Models
Meta emphasizes the importance of diversity in AI. By developing indicators to evaluate geographical disparities and conducting large-scale studies, Meta aims to ensure more inclusive and representative AI-generated content.
