Excited to share these music samples generated from our latest text-to-audio work, AudioLDM. Our paper could be seen on Monday, and more samples of text-to-audio generation, text-guided audio manipulation will be shown soon. AK Welcome to check out our work!

Thank you, AK! The contrastive language-audio pretraining latents enable AudioLDM to learn regenerating audio in training stage, while realizing text to audio in sampling stage. AudioLDM is trained on a single GPU, and advantageous in sample quality and audio manipulation.

Haohe is extremely productive! Try out our AudioLDM if you are interested~~

AudioLDM is on top of the Hugging Face Trending list! 🥳 Glad to see so many of you enjoy this model. BTW, audio style transfer, which is another feature of AudioLDM, will be available on GitHub next week.

Image-to-sound generation with image captioning model and AudioLDM

Update Image to Sound FX — I’ve replaced the gpt2 Image Captioning model by the CoCa: Contrastive Captioners from LAION which seems to be more accurate at describing scenes 😉 — CoCa demo: huggingface.co/spaces/laion/C… Image to SFX demo: huggingface.co/spaces/fffilon…

Very excited to share that our text-to-audio work AudioLDM has been accepted by ICML 2023. It is a great experience to collaborate with Haohe Liu . Finally, we made it! Thank the reviewers and chairs of ICML. Thank AK for sharing our work and videos! ^_^

#AudioLDM #ICML2023 #icml

感谢刘江老师！