Compare AudioLM vs. MuseNet

AudioLM is an innovative audio language model designed to create high-quality, coherent speech and piano music by solely learning from raw audio data, eliminating the need for text transcripts or symbolic forms. It organizes audio in a hierarchical manner through two distinct types of discrete tokens: semantic tokens, which are derived from a self-supervised model to capture both phonetic and melodic structures along with broader context, and acoustic tokens, which come from a neural codec to maintain speaker characteristics and intricate waveform details. This model employs a series of three Transformer stages, initiating with the prediction of semantic tokens to establish the overarching structure, followed by the generation of coarse tokens, and culminating in the production of fine acoustic tokens for detailed audio synthesis. Consequently, AudioLM can take just a few seconds of input audio to generate seamless continuations that effectively preserve voice identity and prosody in speech, as well as melody, harmony, and rhythm in music. Remarkably, evaluations by humans indicate that the synthetic continuations produced are almost indistinguishable from actual recordings, demonstrating the technology's impressive authenticity and reliability. This advancement in audio generation underscores the potential for future applications in entertainment and communication, where realistic sound reproduction is paramount.

We have developed MuseNet, an advanced deep neural network capable of producing 4-minute musical pieces featuring 10 distinct instruments, while seamlessly merging genres ranging from country to the classical compositions of Mozart and even the iconic sounds of the Beatles. Rather than being programmed with musical knowledge, MuseNet identifies and learns patterns of harmony, rhythm, and style through the process of predicting the subsequent token in a vast collection of MIDI files. This innovative model employs the same unsupervised technology as GPT-2, a robust transformer model designed to anticipate the next token in a sequence, whether it pertains to audio or text. Thanks to MuseNet's understanding of diverse musical styles, we are able to create unique blends of musical generations. We eagerly anticipate the creative ways in which both musicians and those without formal training will leverage MuseNet to craft original compositions! Users can select a composer or style and optionally begin with a well-known piece, allowing them to delve into the rich array of musical styles that the model can produce. This opens up exciting possibilities for artistic exploration and experimentation.