AI music generators leave predictable spectral artifacts determined by their architecture. Existing detectors exploit these artifacts with near-perfect accuracy on raw generated tracks, but their performance collapses under simple audio manipulations, such as speed modification or pitch shifting. We address this open robustness problem by introducing a frequency-scaling-invariant detection pipeline that aims to prevent this kind of attack by design. Our method maps audio onto a log-frequency axis via a log-STFT remapping. A single learned cross-correlation filter, combined with max-pooling, provides shift invariance at inference time. Training uses a hybrid loss that jointly supervises binary detection and artifact-peak localization, regularizing boundary weights. Because robustness to speed change is built in by design, the detector is also interpretable: it outputs both a binary decision and an estimate of the applied speed-change factor.

We present a novel method for AI-generated music detection in scenarios where the models that generated the input samples are unknown to the detector (e.g., from a newly released service). Since 2023, there has been a multiplication of user-friendly AI-music generation services (e.g., Suno, Udio), along with regular updates and new features. There is thus a need to address synthetic content detection in an unsupervised way to adapt to this rapidly changing context. This angle has not been much studied in music yet. We propose to study two tasks. First, discriminating between real and synthetic music. This may be approached in a one-class manner, namely, using some baseline real music and trying to determine what falls outside. Second, zero-shot multi-class identification, which is more similar to an unsupervised clustering task on a mix of real and various AI-music generations, where the goal is to create coherent, high-purity clusters. We propose a combination of a previously proposed artifact-extraction method, on top of which we apply non-negative matrix factorization and simple classification and clustering methods. We achieve excellent performance on both tasks, showing that the proposed methods may be used to monitor large-scale catalogs that may receive AI-generated samples from various newly released generative models.

The rapid rise of generative AI has transformed music creation, with millions of users engaging in AI-generated music. Despite its popularity, concerns regarding copyright infringement, job displacement, and ethical implications have led to growing scrutiny and legal challenges. In parallel, AI-detection services have emerged, yet these systems remain largely opaque and privately controlled, mirroring the very issues they aim to address. This paper explores the fundamental properties of synthetic content and how it can be detected. Specifically, we analyze deconvolution modules commonly used in generative models and mathematically prove that their outputs exhibit systematic frequency artifacts -- manifesting as small yet distinctive spectral peaks. This phenomenon, related to the well-known checkerboard artifact, is shown to be inherent to a chosen model architecture rather than a consequence of training data or model weights. We validate our theoretical findings through extensive experiments on open-source models, as well as commercial AI-music generators such as Suno and Udio. We use these insights to propose a simple and interpretable detection criterion for AI-generated music. Despite its simplicity, our method achieves detection accuracy on par with deep learning-based approaches, surpassing 99% accuracy on several scenarios.

In the face of a new era of generative models, the detection of artificially generated content has become a matter of utmost importance. In particular, the ability to create credible minute-long synthetic music in a few seconds on user-friendly platforms poses a real threat of fraud on streaming services and unfair competition to human artists. This paper demonstrates the possibility (and surprising ease) of training classifiers on datasets comprising real audio and artificial reconstructions, achieving a convincing accuracy of 99.8%. To our knowledge, this marks the first publication of a AI-music detector, a tool that will help in the regulation of synthetic media. Nevertheless, informed by decades of literature on forgery detection in other fields, we stress that getting a good test score is not the end of the story. We expose and discuss several facets that could be problematic with such a deployed detector: robustness to audio manipulation, generalisation to unseen models. This second part acts as a position for future research steps in the field and a caveat to a flourishing market of artificial content checkers.

Humans have the tendency to discover and explore. This natural tendency is reflected in data from streaming platforms as the amount of previously unknown content accessed by users. Additionally, in domains such as that of music streaming there is evidence that recommending novel content improves users' experience with the platform. Therefore, understanding users' discovery patterns, such as the amount to which and the way users access previously unknown content, is a topic of relevance for both the scientific community and the streaming industry, particularly the music one. Previous works studied how music consumption differs for users of different traits and looked at diversity, novelty, and consistency over time of users' music preferences. However, very little is known about how users discover and explore previously unknown music, and how this behavior differs for users of varying discovery needs. In this paper we bridge this gap by analyzing data from a survey answered by users of the major music streaming platform Deezer in combination with their streaming data. We first address questions regarding whether users who declare a higher interest in unfamiliar music listen to more diverse music, have more stable music preferences over time, and explore more music within a same time window, compared to those who declare a lower interest. We then investigate which type of music tracks users choose to listen to when they explore unfamiliar music, identifying clear patterns of popularity and genre representativeness that vary for users of different discovery needs. Our findings open up possibilities to infer users' interest in unfamiliar music from streaming data as well as possibilities to develop recommender systems that guide users in exploring music in a more natural way.

In the face of a new era of generative models, the detection of artificially generated content has become a matter of utmost importance. The ability to create credible minute-long music deepfakes in a few seconds on user-friendly platforms poses a real threat of fraud on streaming services and unfair competition to human artists. This paper demonstrates the possibility (and surprising ease) of training classifiers on datasets comprising real audio and fake reconstructions, achieving a convincing accuracy of 99.8 knowledge, this marks the first publication of a music deepfake detector, a tool that will help in the regulation of music forgery. Nevertheless, informed by decades of literature on forgery detection in other fields, we stress that a good test score is not the end of the story. We step back from the straightforward ML framework and expose many facets that could be problematic with such a deployed detector: calibration, robustness to audio manipulation, generalisation to unseen models, interpretability and possibility for recourse. This second part acts as a position for future research steps in the field and a caveat to a flourishing market of fake content checkers.

Recommender systems relying on Language Models (LMs) have gained popularity in assisting users to navigate large catalogs. LMs often exploit item high-level descriptors, i.e. categories or consumption contexts, from training data or user preferences. This has been proven effective in domains like movies or products. However, in the music domain, understanding how effectively LMs utilize song descriptors for natural language-based music recommendation is relatively limited. In this paper, we assess LMs effectiveness in recommending songs based on user natural language descriptions and items with descriptors like genres, moods, and listening contexts. We formulate the recommendation task as a dense retrieval problem and assess LMs as they become increasingly familiar with data pertinent to the task and domain. Our findings reveal improved performance as LMs are fine-tuned for general language similarity, information retrieval, and mapping longer descriptions to shorter, high-level descriptors in music.

The truncated singular value decomposition is a widely used methodology in music recommendation for direct similar-item retrieval and downstream tasks embedding musical items. This paper investigates a curious effect that we show naturally occurring on many recommendation datasets: spiking formations in the embedding space. We first propose a metric to quantify this spiking organization’s strength, then mathematically prove its origin tied to underlying communities of items of varying internal popularity. With this new-found theoretical understanding, we finally open the topic with an industrial use case of estimating how music embeddings’ top-k similar items will change over time under the addition of data.

Systèmes de recommandation musicale interprétables « Pourquoi est-ce qu’on me recommande toujours les même musiques ? » « Pourquoi notre système recommande-t’il cela aux utilisateurs ? » De nos jours, les plateformes de streaming sont le moyen le plus courant d'écouter de la musique enregistrée. Pourtant, les recommandations musicales — au cœur de ces plateformes — sont loin d’être une mince affaire. Il arrive parfois qu’utilisateurs et ingénieurs soient tout aussi perplexes du comportement d’un système de recommandation musicale (SRM). Les SRM ont été utilisés avec succès pour aider à explorer des catalogues comptant des dizaines de millions de titres musicaux. Construits et optimisés pour la précision, les SRM industriels sont souvent assez complexes. Ils peuvent en outre dépendre de nombreux modules interconnectés qui, notamment, analysent les signaux audio, récupèrent les métadonnées d’albums et artistes et les interactions des utilisateurs du service, et estiment des similarités basées sur du filtrage collaboratif. Cette complexité va en l’encontre de la capacité d'expliquer les recommandations et, plus généralement, ces systèmes. Pourtant, les explications sont essentielles pour fidéliser des utilisateurs sur le long termes avec un système qu'ils peuvent comprendre (et pardonner), et pour les propriétaires du système pour rationaliser les erreurs dudit système. L'interprétabilité peut également être nécessaire pour vérifier l'équité d'une décision ou peut être envisagées comme un moyen de rendre les recommandations plus contrôlables. Nous pouvons également récursivement demander : pourquoi une méthode d'explication explique-t-elle d'une certaine manière ? Cette explication est-elle pertinente ? Quelle pourrait être une meilleure explication ? Toutes ces questions sont liées à l'interprétabilité des SRM. Dans une première partie, nous explorons les multiples visages de l'interprétabilité dans diverses tâches de recommandation. En effet, puisqu'il n'y a pas une seule tâche de recommandation mais plusieurs (e.g., recommandation séquentielle, continuation de playlists, similarité artistes), ainsi que de nombreuses modalités de représentation de la musique (e.g., métadonnées, signaux audio, plongements), il y a autant de tâches possibles d’explications nécessitant des ajustements. Notre étude a été guidée par l’exploration des modalités sus-mentionnées : l'interprétation des signaux implicites utilisateurs, des caractéristiques, des signaux audio, et des inter-similarités. Notre thèse présente plusieurs nouvelles méthodes pour l'IA explicable (XAI) et plusieurs résultats théoriques, portant un nouvel éclairage sur notre compréhension des méthodes passées. Néanmoins, les méthodes d’explications peuvent à leur tour manquer d'interprétabilité. C'est pourquoi, une deuxième partie, nous avons jugé essentiel de prendre du recul par rapport aux discours habituels de l’IA et d'essayer de répondre à une question paradoxalement peu claire pour l’XAI : « Qu'est-ce que l'interprétabilité ? » En s'appuyant sur des concepts issus des sciences sociales, nous soulignons qu'il existe un décalage entre la manière dont les explications de l'XAI sont générées et la manière dont les humains expliquent réellement. Nous suggérons que la recherche actuelle a tendance à trop s'appuyer sur des intuitions et des réductions hâtive de réalités complexes en termes mathématiques commodes, conduisant à ériger des hypothèses en normes discutables (e.g., la parcimonie entraîne l'interprétabilité). Nous avons pensé cette partie comme un tutoriel destiné aux chercheurs en IA afin de renforcer leur connaissance des explications avec un vocabulaire précis et une perspective plus large. Nous résumons des conseils pratiques et mettons en évidence des branches moins populaires de l'XAI mieux alignées avec l’humain. Cela nous permet de formuler une perspective globale pour notre domaine de l'XAI, y compris ses prochaines étapes les plus critiques et prometteuses ainsi que ses lacunes à surmonter.

The most common way to listen to recorded music nowadays is via streaming platforms which provide access to tens of millions of tracks. To assist users in effectively browsing these large catalogs, the integration of Music Recommender Systems (MRSs) has become essential. Current real-world MRSs are often quite complex and optimized for recommendation accuracy. They combine several building blocks based on collaborative filtering and content-based recommendation. This complexity can hinder the ability to explain recommendations to end users, which is particularly important for recommendations perceived as unexpected or inappropriate. While pure recommendation performance often correlates with user satisfaction, explainability has a positive impact on other factors such as trust and forgiveness, which are ultimately essential to maintain user loyalty. In this article, we discuss how explainability can be addressed in the context of MRSs. We provide perspectives on how explainability could improve music recommendation algorithms and enhance user experience. First, we review common dimensions and goals of recommenders' explainability and in general of eXplainable Artificial Intelligence (XAI), and elaborate on the extent to which these apply -- or need to be adapted -- to the specific characteristics of music consumption and recommendation. Then, we show how explainability components can be integrated within a MRS and in what form explanations can be provided. Since the evaluation of explanation quality is decoupled from pure accuracy-based evaluation criteria, we also discuss requirements and strategies for evaluating explanations of music recommendations. Finally, we describe the current challenges for introducing explainability within a large-scale industrial music recommender system and provide research perspectives.

Music signals are difficult to interpret from their low-level features, perhaps even more than images: e.g. highlighting part of a spectrogram or an image is often insufficient to convey high-level ideas that are genuinely relevant to humans. In computer vision, concept learning was therein proposed to adjust explanations to the right abstraction level (e.g. detect clinical concepts from radiographs). These methods have yet to be used for MIR. In this paper, we adapt concept learning to the realm of music, with its particularities. For instance, music concepts are typically non-independent and of mixed nature (e.g. genre, instruments, mood), unlike previous work that assumed disentangled concepts. We propose a method to learn numerous music concepts from audio and then automatically hierarchise them to expose their mutual relationships. We conduct experiments on datasets of playlists from a music streaming service, serving as a few annotated examples for diverse concepts. Evaluations show that the mined hierarchies are aligned with both ground-truth hierarchies of concepts -- when available -- and with proxy sources of concept similarity in the general case.

Feature attribution is often loosely presented as the process of selecting a subset of relevant features as a rationale of a prediction. Task-dependent by nature, precise definitions of "relevance" encountered in the literature are however not always consistent. This lack of clarity stems from the fact that we usually do not have access to any notion of ground-truth attribution and from a more general debate on what good interpretations are. In this paper we propose to formalise feature selection/attribution based on the concept of relaxed functional dependence. In particular, we extend our notions to the instance-wise setting and derive necessary properties for candidate selection solutions, while leaving room for task-dependence. By computing ground-truth attributions on synthetic datasets, we evaluate many state-of-the-art attribution methods and show that, even when optimised, some fail to verify the proposed properties and provide wrong solutions.

Explaining recommendations enables users to understand whether recommended items are relevant to their needs and has been shown to increase their trust in the system. More generally, if designing explainable machine learning models is key to check the sanity and robustness of a decision process and improve their efficiency, it however remains a challenge for complex architectures, especially deep neural networks that are often deemed "black-box". In this paper, we propose a novel formulation of interpretable deep neural networks for the attribution task. Differently to popular post-hoc methods, our approach is interpretable by design. Using masked weights, hidden features can be deeply attributed, split into several input-restricted sub-networks and trained as a boosted mixture of experts. Experimental results on synthetic data and real-world recommendation tasks demonstrate that our method enables to build models achieving close predictive performances to their non-interpretable counterparts, while providing informative attribution interpretations.

This paper presents a method to automatically and efficiently detect face tampering in videos, and particularly focuses on two recent techniques used to generate hyper-realistic forged videos: Deepfake and Face2Face. Traditional image forensics techniques are usually not well suited to videos due to the compression that strongly degrades the data. Thus, this paper follows a deep learning approach and presents two networks, both with a low number of layers to focus on the mesoscopic properties of images. We evaluate those fast networks on both an existing dataset and a dataset we have constituted from online videos. The tests demonstrate a very successful detection rate with more than 98% for Deepfake and 95% for Face2Face.