We present and release MIDI-GPT, a generative system based on the Transformer architecture that is designed for computer-assisted music composition workflows. MIDI-GPT supports the infilling of musical material at the track and bar level, and can condition generation on attributes including: instrument type, musical style, note density, polyphony level, and note duration. In order to integrate these features, we employ an alternative representation for musical material, creating a time-ordered sequence of musical events for each track and concatenating several tracks into a single sequence, rather than using a single time-ordered sequence where the musical events corresponding to different tracks are interleaved. We also propose a variation of our representation allowing for expressiveness. We present experimental results that demonstrate that MIDI-GPT is able to consistently avoid duplicating the musical material it was trained on, generate music that is stylistically similar to the training dataset, and that attribute controls allow enforcing various constraints on the generated material. We also outline several real-world applications of MIDI-GPT, including collaborations with industry partners that explore the integration and evaluation of MIDI-GPT into commercial products, as well as several artistic works produced using it.

Lyrics-to-melody generation is an interesting and challenging topic in AI music research field. Due to the difficulty of learning the correlations between lyrics and melody, previous methods suffer from low generation quality and lack of controllability. Controllability of generative models enables human interaction with models to generate desired contents, which is especially important in music generation tasks towards human-centered AI that can facilitate musicians in creative activities. To address these issues, we propose a controllable lyrics-to-melody generation network, ConL2M, which is able to generate realistic melodies from lyrics in user-desired musical style. Our work contains three main novelties: 1) To model the dependencies of music attributes cross multiple sequences, inter-branch memory fusion (Memofu) is proposed to enable information flow between multi-branch stacked LSTM architecture; 2) Reference style embedding (RSE) is proposed to improve the quality of generation as well as control the musical style of generated melodies; 3) Sequence-level statistical loss (SeqLoss) is proposed to help the model learn sequence-level features of melodies given lyrics. Verified by evaluation metrics for music quality and controllability, initial study of controllable lyrics-to-melody generation shows better generation quality and the feasibility of interacting with users to generate the melodies in desired musical styles when given lyrics.

Sony's Computer Science Laboratory in Paris has been working on the research and development of pioneering music technologies since 1997, and a blog post from one of the lab's teams this week unveiled an advancement that could reverberate throughout the music world. The team's Flow Machines project successfully created two entire pop songs composed by artificial intelligence, after learning musical styles from a massive database. After "exploiting unique combinations of style transfer, optimization and interaction techniques, it can compose in any style," the post reads. The project's success aligns with the team's goals, which, according to the lab's site, has the aim to "abstract'style' from concrete corpora (text, music, etc.), and turn it into a malleable substance that acts as a texture." Though the team has been successful in the past with constraint-based spatialization -- intelligent music scheduling using metadata and award-winning systems (MusicSpace, PathBuilder, Virtuoso, etc.) -- the work it showed off this week might take home the prize for being one of the most impressive.

The Beatles may have disbanded decades ago, but thanks to artificial intelligence, the group is being reanimated -- sort of. Researchers at Sony are at work on an algorithm capable of generating new songs based on iconic musical styles, and their first crack at putting it in action starts with the iconic rock band. The song is called "Daddy's Car," which bears an eerie resemblance to the many songs that John, Paul, George, and Ringo used to play. It was made using a system called FlowMachines developed by the team at Sony CSL Research Laboratory, and was trained on a huge database of 13,000 songs. Once it understands a desired style, it can, with the help of a producer/arranger, churn out new songs in a variety of musical styles.

Creating a musical fitness function is largely subjective and can be critically affected by the designer's biases. Previous attempts to create such functions for use in genetic algorithms lack scope or are prejudiced to a certain genre of music. They also are limited to producing music strictly in the style determined by the programmer. We show in this paper that musical feature extractors, which avoid the challenges of qualitative judgment, enable creation of a multi-objective function for direct music production. The main result is that the multi-objective fitness function enables creation of music with varying identifiable styles. To demonstrate this, we use three different multi-objective fitness functions to create three distinct sets of musical melodies. We then evaluate the distinctness of these sets using three different approaches: a set of traditional computational clustering metrics; a survey of non-musicians; and analysis by three trained musicians.