Vincent Lostanlen – Page 2

Learning to Solve Inverse Problems for Perceptual Sound Matching @ IEEE TASLP

Perceptual sound matching (PSM) aims to find the input parameters to a synthesizer so as to best imitate an audio target. Deep learning for PSM optimizes a neural network to analyze and reconstruct prerecorded samples. In this context, our article addresses the problem of designing a suitable loss function when the training set is generated by a differentiable synthesizer. Our main contribution is perceptual–neural–physical loss (PNP), which aims at addressing a tradeoff between perceptual relevance and computational efficiency. The key idea behind PNP is to linearize the effect of synthesis parameters upon auditory features in the vicinity of each training sample. The linearization procedure is massively parallelizable, can be precomputed, and offers a 100-fold speedup during gradient descent compared to differentiable digital signal processing (DDSP). We show that PNP is able to accelerate DDSP with joint time–frequency scattering transform (JTFS) as auditory feature while preserving its perceptual fidelity. Additionally, we evaluate the impact of other design choices in PSM: parameter rescaling, pretraining, auditory representation, and gradient clipping. We report state-of-the-art results on both datasets and find that PNP-accelerated JTFS has greater influence on PSM performance than any other design choice.

Model-Based Deep Learning for Music Information Research @ IEEE SPM

In this article, we investigate the notion of model-based deep learning in the realm of music information research (MIR). Loosely speaking, we refer to the term model-based deep learning for approaches that combine traditional knowledge-based methods with data-driven techniques, especially those based on deep learning, within a diff erentiable computing framework. In music, prior knowledge for instance related to sound production, music perception or music composition theory can be incorporated into the design of neural networks and associated loss functions. We outline three specific scenarios to illustrate the application of model-based deep learning in MIR, demonstrating the implementation of such concepts and their potential.

BioacAI doctoral network workshop in Czech Republic

From May 6th to May 10th, 2024, PhD student Yasmine Benhamadi and CNRS scientist Vincent Lostanlen have attended the first internal workshop of the BioacAI doctoral network. The Czech University of Life Sciences in Prague hosted the event in its University Forest Establishment, an ancient castle in the town of Kostelec nad Černými lesy.

Action “Musiscale” au symposium du GDR MaDICS

Le 30 mai 2024 à Blois, se tenait le sixième symposium du GDR MaDICS : masses de données, informations et connaissances en sciences. Dans le cadre de l’action “Musiscale : modélisation multi-échelles de masses de données musicales”, j’ai présenté les travaux de l’équipe sur la diffusion en ondelettes (scattering transform) ainsi que sur les réseaux de neurones multirésolution (MuReNN pour multi-resolution neural networks).

Japanese–French Frontiers of Science Symposium 「日仏先端科学シンポジウム」

Towards multisensory control of physical modeling synthesis @ Inter-Noise

Physical models of musical instruments offer an interesting tradeoff between computational efficiency and perceptual fidelity. Yet, they depend on a multidimensional space of user-defined parameters whose exploration by trial and error is impractical. Our article addresses this issue by combining two ideas: query by example and gestural control. On one hand, we train a deep… Continue reading Towards multisensory control of physical modeling synthesis @ Inter-Noise

11 juin 2024 : colloque “Capteurs acoustiques en environnement” à Nantes

Le son, en tant que vecteur d’information, est une aubaine pour les sciences naturelles. À l’heure des smartphones et de l’Internet des objets, il devient possible de décrire dans le détail les propriétés acoustiques d’un environnement, que celui-ci soit naturel ou industrialisé. Des algorithmes d’intelligence artificielle (IA) sont alors requis pour traiter automatiquement les données massives ainsi collectées et interagir utilement avec l’humain. Mais en pratique, un tel programme de recherche soulève des problèmes de fiabilité, de durabilité, de sécurité informatique et de mesure de l’incertitude.

Structure Versus Randomness in Computer Music and the Scientific Legacy of Jean-Claude Risset @ JIM

According to Jean-Claude Risset (1938–2016), “art and science bring about complementary kinds of knowledge”. In 1969, he presented his piece Mutations as “[attempting] to explore […] some of the possibilities offered by the computer to compose at the very level of sound—to compose sound itself, so to speak.” In this article, I propose to take the same motto as a starting point, yet while adopting a mathematical and technological outlook, more so than a musicological one.

Instabilities in Convnets for Raw Audio @ IEEE SPL

What makes waveform-based deep learning so hard? Despite numerous attempts at training convolutional neural networks (convnets) for filterbank design, they often fail to outperform hand-crafted baselines. These baselines are linear time-invariant systems: as such, they can be approximated by convnets with wide receptive fields. Yet, in practice, gradient-based optimization leads to suboptimal approximations. In our… Continue reading Instabilities in Convnets for Raw Audio @ IEEE SPL

PhD offer: Machine learning on solar-powered environmental sensors

Many biological and geophysical phenomena follow a near-periodic day-night cycle, known as circadian rhythm. When designing AI-enabled autonomous sensors for environmental modeling, this circadian rhythm poses both a challenge and an opportunity.