Models for audio generation are typically trained on hours of recordings. Here, we illustrate that capturing the essence of an audio source is typically possible from as little as a few tens of seconds from a single training signal. Specifically, we present a GAN-based generative model that can be trained on one short audio signal from any domain (e.g.

How does textual representation of audio relate to the Large Language Model's (LLMs) learning about the audio world? This research investigates the extent to which LLMs can be prompted to generate audio, despite their primary training in textual data. We employ a three-tier approach, progressively increasing the complexity of audio generation: 1) Musical Notes, 2) Environmental Sounds, and 3) Human Speech. To bridge the gap between text and audio, we leverage code as an intermediary, prompting LLMs to generate code that, when executed, produces the desired audio output. To evaluate the quality and accuracy of the generated audio, we employ FAD and CLAP scores. Our findings reveal that while LLMs can generate basic audio features, their performance deteriorates as the complexity of the audio increases. This suggests that while LLMs possess a latent understanding of the auditory world, their ability to translate this understanding into tangible audio output remains rudimentary. Further research into techniques that can enhance the quality and diversity of LLM-generated audio can lead to an improvement in the performance of text-based LLMs in generating audio.

Models for audio generation are typically trained on hours of recordings. Here, we illustrate that capturing the essence of an audio source is typically possible from as little as a few tens of seconds from a single training signal. Specifically, we present a GAN-based generative model that can be trained on one short audio signal from any domain (e.g. Once trained, our model can generate random samples of arbitrary duration that maintain semantic similarity to the training waveform, yet exhibit new compositions of its audio primitives. This enables a long line of interesting applications, including generating new jazz improvisations or new a-cappella rap variants based on a single short example, producing coherent modifications to famous songs (e.g. We show that in all cases, no more than 20 seconds of training audio commonly suffice for our model to achieve state-of-the-art results.

We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to-video synthesis, video personalization, video editing, video-to-audio generation, and text-to-audio generation. Our largest video generation model is a 30B parameter transformer trained with a maximum context length of 73K video tokens, corresponding to a generated video of 16 seconds at 16 frames-per-second. We show multiple technical innovations and simplifications on the architecture, latent spaces, training objectives and recipes, data curation, evaluation protocols, parallelization techniques, and inference optimizations that allow us to reap the benefits of scaling pre-training data, model size, and training compute for training large scale media generation models. We hope this paper helps the research community to accelerate progress and innovation in media generation models. All videos from this paper are available at https://go.fb.me/MovieGenResearchVideos.

In video game design, audio (both environmental background music and object sound effects) play a critical role. Sounds are typically pre-created assets designed for specific locations or objects in a game. However, user-generated content is becoming increasingly popular in modern games (e.g. building custom environments or crafting unique objects). Since the possibilities are virtually limitless, it is impossible for game creators to pre-create audio for user-generated content. We explore the use of generative artificial intelligence to create music and sound effects on-the-fly based on user-generated content. We investigate two avenues for audio generation: 1) text-to-audio: using a text description of user-generated content as input to the audio generator, and 2) image-to-audio: using a rendering of the created environment or object as input to an image-to-text generator, then piping the resulting text description into the audio generator. In this paper we discuss ethical implications of using generative artificial intelligence for user-generated content and highlight two prototype games where audio is generated for user-created environments and objects.

This article is an installment of Future Explored, a weekly guide to world-changing technology. You can get stories like this one straight to your inbox every Thursday morning by subscribing here. In January, Microsoft unveiled an AI that can clone a speaker's voice after hearing them talk for just three seconds. While this system, VALL-E, was far from the first voice cloning AI, its accuracy and need for such a small audio sample set a new bar for the tech. Microsoft has now raised that bar again with an update called "VALL-E X," which can clone a voice from a short sample (4 to 10 seconds) and then use it to synthesize speech in a different language, all while preserving the original speaker's voice, emotion, and tone.

As the title suggests, in this blog we are going to learn about text to speech (TTS) synthesis. What is the first bell which rings in your mind when you listen to text to speech? For me, it's Alexa, Google Home, Siri, and many other conversational bots that are on an exponential rise currently. Advances in deep learning research have helped us to generate human-like voices, so let's see how we can use that. I'll start with a few definitions, but if you want to understand these more then read this blog first.