Agents engaged in creative joint actions might need to find a balance between the demands of doing something collectively, by adopting congruent and interacting behaviors, and the goal of delivering a creative output, which can eventually benefit from disagreements and autonomous behaviors. Here, we investigate this idea in the context of collective free improvisation -a paradigmatic example of group creativity in which musicians aim at creating music that is as complex and unprecedented as possible without relying on predefined plans or individual roles.Controlling for both the familiarity between the musicians and their physical co-presence, duos of improvisers were asked to freely improvise together and to individually annotate their performances with a digital interface, indicating at each time whether they were playing "with", "against", or "without" their partner. At an individual level, we found that musicians largely intended to converge with their co-improviser, making only occasional use of non-cooperative or non-interactive modes such as "playing against" or "playing without". By contrast, at the group level, musicians tended to combine their relational intents in such a way as to create interactional dissensus. We also demonstrate that co-presence and familiarity act as interactional smoothers: they increase the agents' overall level of relational plasticity and allow for the exploration of less cooperative behaviors. Overall, our findings suggest that relational intents might function as a primary resource for creative joint actions.
This paper studies the prediction of chord progressions for jazz music by relying on machine learning models. The motivation of our study comes from the recent success of neural networks for performing automatic music composition. Although high accuracies are obtained in single-step prediction scenarios, most models fail to generate accurate multistep chord predictions. In this paper, we postulate that this comes from the multi-scale structure of musical information and propose new architectures based on an iterative temporal aggregation of input labels. Specifically, the input and ground truth labels are merged into increasingly large temporal bags, on which we train a family of encoder-decoder networks for each temporal scale. In a second step, we use these pretrained encoder bottleneck features at each scale in order to train a final encoder-decoder network. Furthermore, we rely on different reductions of the initial chord alphabet into three adapted chord alphabets. We perform evaluations against several state-of-the-art models and show that our multi-scale architecture outperforms existing methods in terms of accuracy and perplexity, while requiring relatively few parameters. We analyze musical properties of the results, showing the influence of downbeat position within the analysis window on accuracy, and evaluate errors using a musically-informed distance metric.
This article focuses on the introduction of control, authoring, and composition in human-computer music improvisation through the description of a guided music generation model and a reactive architecture, both implemented in the software ImproteK. This interactive music system is used with expert improvisers in work sessions and performances of idiomatic and pulsed music and more broadly in situations of structured or composed improvisation. The article deals with the integration of temporal specifications in the music generation process by means of a fixed or dynamic "scenario" and addresses the issue of the dialectic between reactivity and planning in interactive music improvisation. It covers the different levels involved in machine improvisation: the integration of anticipation relative to a predefined structure in a guided generation process at a symbolic level, an architecture combining this anticipation with reactivity using mixed static/dynamic scheduling techniques, and an audio rendering module performing live re-injection of captured material in synchrony with a non-metronomic beat. Finally, it sketches a framework to compose improvisation sessions at the scenario level, extending the initial musical scope of the system. All of these points are illustrated by videos of performances or work sessions with musicians.
To cite this version:Jérôme Nika, Marc Chemillier. Improvisation musicale homme-machine guidée par un scénario temporel. Technique et Science Informatiques, Hermès-Lavoisier, 2015 Cet article propose un modèle pour l'improvisation musicale guidée par une structure formalisée. Il exploite les connaissances a priori du contexte d'improvisation pour introduire de l'anticipation dans le processus de génération. « Improviser » signifie ici articuler une mémoire musicale annotée avec un « scénario » guidant l'improvisation (par exemple une progression harmonique dans le cas du jazz). Le scénario et la séquence étiquetant la mémoire sont représentés comme des mots construits sur un même alphabet. Cet alphabet définit les classes d'équivalences choisies pour décrire les éléments musicaux constituant la mémoire. La navigation dans cette mémoire assure la cohérence et l'homogénéité du résultat musical en exploitant les motifs communs aux deux séquences tout en étant capable de s'éloigner du matériau musical d'origine. Le modèle et l'architecture d'ordonnancement décrits dans cet article sont implémen-tés dans le système d'improvisation ImproteK, utilisé à plusieurs reprises avec des musiciens experts.ABSTRACT. This paper presents a computer model for musical improvisation guided by a formalized structure. It uses the prior knowledge of the temporal structure of the improvisation to introduce some anticipation in the music generation process. In this framework "improvising" amounts to articulating an annotated memory and a "scenario" guiding and constraining the improvisation (for example a given chord progression in the case of jazz improvisation). The scenario and the sequence describing the musical memory are represented as words defined over the same alphabet. This alphabet describes the equivalence classes chosen to label the musical contents of the online or offline memory. The navigation through this memory searches for continuity in the musical discourse by exploiting similar patterns in the sequences, as well as the ability to go beyond the simple copy of their associated musical contents. The model and the scheduling architecture described in this paper are implemented in the improvisation software ImproteK which has been used at several occasions with expert musicians.1 MOTS-CLÉS : improvisation musicale guidée, scénario, recherche de motifs, recherche de préfixes, indexation, apprentissage, interaction, ImproteK.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.