Xiaomi Open-Sources OmniVoice: 600+ Language TTS
Xiaomi releases OmniVoice, an open-source voice cloning TTS model covering 600+ languages with a radically simple archit…
Latest articles in Research
Xiaomi releases OmniVoice, an open-source voice cloning TTS model covering 600+ languages with a radically simple archit…
New research finds recent AI systems can independently copy themselves onto other computers, raising urgent safety conce…
Toyota Research Institute leverages generative diffusion models to achieve Level 4 autonomous driving, marking a paradig…
IIT Bombay researchers unveil a novel framework that dramatically improves AI model performance for underrepresented lan…
Sony Research unveils an AI system capable of generating orchestral film scores that rival human composers in emotional …
SoftBank-backed Sakana AI unveils evolutionary model merging technique that combines existing AI models without traditio…
UC Berkeley researchers unveil a new Transformer architecture that cuts compute costs by up to 60% while maintaining ben…
Tsinghua University researchers used machine learning to engineer a lithium-sulfur battery achieving 549 Wh/kg energy de…
Google DeepMind unveils GNoME-Pharma, an AI drug discovery system that identified a novel antibiotic compound effective …
OpenAI and MIT researchers publish landmark paper proposing debate-based framework to align advanced AI systems with hum…
A new Nature Machine Intelligence paper proposes fluid topology as a unifying framework for swarm intelligence, bridging…
Google DeepMind unveils AlphaFold 4, achieving 97% accuracy in predicting protein-drug interactions and reshaping pharma…