Today's featured AI news highlights, enjoy reading 👇

🎧 Eleven Labs launches AI podcast feature, supporting text-to-podcast conversion in 32 languages.

🎨 IC-Light updates to v2, enhancing lighting effects and editing capabilities, available for free on Hugging Face but not yet open-source.

📹 Google releases CATD model, capable of converting single-view videos into multi-view perspectives and building dynamic 4D scenes, significant for film production and VR/AR fields.

Latest News

1. Eleven Labs launches AI podcast feature.

Now you can generate AI personal podcasts in 32 languages from any PDF, article, e-book, link, or text through the ElevenReander iOS client.

Detailed introduction: https://elevenlabs.io/genfm

However, Eleven Labs' Chinese has always been relatively weak. Personally, I think Feynman's Chinese accent is slightly better, you can give it a try.

AI Art

1. IC-Light v2 version has been updated.

It can generate stronger lighting effects and allow modifications, though it's not open-source yet.

Try it online: https://huggingface.co/spaces/lllyasviel/iclight-v2-vary

Currently available for testing on Hugging Face, with impressive results. Worth checking out if you're interested.

Cutting-edge Technology

1. Google introduces multi-view video generation model: CATD.

Using this model, it's possible to generate multiple viewpoint videos from a single-view video and construct dynamic 4D scenes from these videos.

Detailed introduction: https://cat-4d.github.io/

Additionally, it can process and manipulate camera movements and object motions in videos, which is highly beneficial for filmmaking, virtual reality, and augmented reality.