The ICASSP 2026 Automatic Song Aesthetics Evaluation Challenge

This paper summarizes the ICASSP 2026 Automatic Song Aesthetics Evaluation (ASAE) Challenge, which focuses on predicting the subjective aesthetic scores of AI-generated songs. The challenge consists of two tracks: Track 1 targets the prediction of the overall musicality score, while Track 2 focuses on predicting five fine-grained aesthetic scores. The challenge attracted strong interest from the research community and received numerous submissions from both academia and industry. Top-performing systems significantly surpassed the official baseline, demonstrating substantial progress in aligning objective metrics with human aesthetic preferences. The outcomes establish a standardized benchmark and advance human-aligned evaluation methodologies for modern music generation systems.

Semantic visually-guided acoustic highlighting with large vision-language models

Balancing dialogue, music, and sound effects with accompanying video is crucial for immersive storytelling, yet current audio mixing workflows remain largely manual and labor-intensive. While recent advancements have introduced the visually guided acoustic highlighting task, which implicitly rebalances audio sources using multimodal guidance, it remains unclear which visual aspects are most effective as conditioning signals.We address this gap through a systematic study of whether deep video understanding improves audio remixing. Using textual descriptions as a proxy for visual analysis, we prompt large vision-language models to extract six types of visual-semantic aspects, including object and character appearance, emotion, camera focus, tone, scene background, and inferred sound-related cues. Through extensive experiments, camera focus, tone, and scene background consistently yield the largest improvements in perceptual mix quality over state-of-the-art baselines. Our findings (i) identify which visual-semantic cues most strongly support coherent and visually aligned audio remixing, and (ii) outline a practical path toward automating cinema-grade sound design using lightweight guidance derived from large vision-language models.

Elastic overtones: an equal temperament 12 tone music system with "perfect" fifths

The impossibility of a transposable 12 semitone tuning of the octave arises from the mathematical fact that $2 \times 2^{7/12} \neq 3$ i.e., the second harmonic of the fifth can not exactly match the third harmonic of the fundamental. This in turn, stems from the whole number harmonic structure of western music, and the subsequent fundamental character of the octave interval as multiples of 2 in frequency, a property inherited by our music system from the physics of instruments with vibrating elements being to a good approximation one dimensional. In the current era of electronic music, one can relax the above assumptions to construct an analogous music system where all the structural properties of the standard music system are preserved, but where harmonics are not whole number multiples of the fundamental frequency, and the octave is no longer a factor of 2 in frequency. This now allows to construct a transposable 12 semitone music system where the second harmonic of the fifth exactly matches the third harmonic of the fundamental. The enhanced harmonic qualities of this system recover to a good approximation the musical qualities of Just Intonation, whilst retaining by construction all the versatility and modulating ability of 12TET.

MoE Adapter for Large Audio Language Models: Sparsity, Disentanglement, and Gradient-Conflict-Free

Extending the input modality of Large Language Models~(LLMs) to the audio domain is essential for achieving comprehensive multimodal perception. However, it is well-known that acoustic information is intrinsically \textit{heterogeneous}, entangling attributes such as speech, music, and environmental context. Existing research is limited to a dense, parameter-shared adapter to model these diverse patterns, which induces \textit{gradient conflict} during optimization, as parameter updates required for distinct attributes contradict each other. To address this limitation, we introduce the \textit{\textbf{MoE-Adapter}}, a sparse Mixture-of-Experts~(MoE) architecture designed to decouple acoustic information. Specifically, it employs a dynamic gating mechanism that routes audio tokens to specialized experts capturing complementary feature subspaces while retaining shared experts for global context, thereby mitigating gradient conflicts and enabling fine-grained feature learning. Comprehensive experiments show that the MoE-Adapter achieves superior performance on both audio semantic and paralinguistic tasks, consistently outperforming dense linear baselines with comparable computational costs. Furthermore, we will release the related code and models to facilitate future research.

Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Control

Recent commercial systems such as Suno demonstrate strong capabilities in long-form song generation, while academic research remains largely non-reproducible due to the lack of publicly available training data, hindering fair comparison and progress. To this end, we release a fully open-source system for long-form song generation with fine-grained style conditioning, including a licensed synthetic dataset, training and evaluation pipelines, and Muse, an easy-to-deploy song generation model. The dataset consists of 116k fully licensed synthetic songs with automatically generated lyrics and style descriptions paired with audio synthesized by SunoV5. We train Muse via single-stage supervised finetuning of a Qwen-based language model extended with discrete audio tokens using MuCodec, without task-specific losses, auxiliary objectives, or additional architectural components. Our evaluations find that although Muse is trained with a modest data scale and model size, it achieves competitive performance on phoneme error rate, text--music style similarity, and audio aesthetic quality, while enabling controllable segment-level generation across different musical structures. All data, model weights, and training and evaluation pipelines will be publicly released, paving the way for continued progress in controllable long-form song generation research. The project repository is available at https://github.com/yuhui1038/Muse.

Density Matrix RNN (DM-RNN): A Quantum Information Theoretic Framework for Modeling Musical Context and Polyphony

Classical Recurrent Neural Networks (RNNs) summarize musical context into a deterministic hidden state vector, imposing an information bottleneck that fails to capture the inherent ambiguity in music. We propose the Density Matrix RNN (DM-RNN), a novel theoretical architecture utilizing the Density Matrix. This allows the model to maintain a statistical ensemble of musical interpretations (a mixed state), capturing both classical probabilities and quantum coherences. We rigorously define the temporal dynamics using Quantum Channels (CPTP maps). Crucially, we detail a parameterization strategy based on the Choi-Jamiolkowski isomorphism, ensuring the learned dynamics remain physically valid (CPTP) by construction. We introduce an analytical framework using Von Neumann Entropy to quantify musical uncertainty and Quantum Mutual Information (QMI) to measure entanglement between voices. The DM-RNN provides a mathematically rigorous framework for modeling complex, ambiguous musical structures.

Summary of The Inaugural Music Source Restoration Challenge

Music Source Restoration (MSR) aims to recover original, unprocessed instrument stems from professionally mixed and degraded audio, requiring the reversal of both production effects and real-world degradations. We present the inaugural MSR Challenge, which features objective evaluation on studio-produced mixtures using Multi-Mel-SNR, Zimtohrli, and FAD-CLAP, alongside subjective evaluation on real-world degraded recordings. Five teams participated in the challenge. The winning system achieved 4.46 dB Multi-Mel-SNR and 3.47 MOS-Overall, corresponding to relative improvements of 91% and 18% over the second-place system, respectively. Per-stem analysis reveals substantial variation in restoration difficulty across instruments, with bass averaging 4.59 dB across all teams, while percussion averages only 0.29 dB. The dataset, evaluation protocols, and baselines are available at https://msrchallenge.com/.

Learning from Limited Labels: Transductive Graph Label Propagation for Indian Music Analysis

Supervised machine learning frameworks rely on extensive labeled datasets for robust performance on real-world tasks. However, there is a lack of large annotated datasets in audio and music domains, as annotating such recordings is resource-intensive, laborious, and often require expert domain knowledge. In this work, we explore the use of label propagation (LP), a graph-based semi-supervised learning technique, for automatically labeling the unlabeled set in an unsupervised manner. By constructing a similarity graph over audio embeddings, we propagate limited label information from a small annotated subset to a larger unlabeled corpus in a transductive, semi-supervised setting. We apply this method to two tasks in Indian Art Music (IAM): Raga identification and Instrument classification. For both these tasks, we integrate multiple public datasets along with additional recordings we acquire from Prasar Bharati Archives to perform LP. Our experiments demonstrate that LP significantly reduces labeling overhead and produces higher-quality annotations compared to conventional baseline methods, including those based on pretrained inductive models. These results highlight the potential of graph-based semi-supervised learning to democratize data annotation and accelerate progress in music information retrieval.

Mathematical Foundations of Polyphonic Music Generation via Structural Inductive Bias

This monograph introduces a novel approach to polyphonic music generation by addressing the "Missing Middle" problem through structural inductive bias. Focusing on Beethoven's piano sonatas as a case study, we empirically verify the independence of pitch and hand attributes using normalized mutual information (NMI=0.167) and propose the Smart Embedding architecture, achieving a 48.30% reduction in parameters. We provide rigorous mathematical proofs using information theory (negligible loss bounded at 0.153 bits), Rademacher complexity (28.09% tighter generalization bound), and category theory to demonstrate improved stability and generalization. Empirical results show a 9.47% reduction in validation loss, confirmed by SVD analysis and an expert listening study (N=53). This dual theoretical and applied framework bridges gaps in AI music generation, offering verifiable insights for mathematically grounded deep learning.

From Score to Sound: An End-to-End MIDI-to-Motion Pipeline for Robotic Cello Performance

Robot musicians require precise control to obtain proper note accuracy, sound quality, and musical expression. Performance of string instruments, such as violin and cello, presents a significant challenge due to the precise control required over bow angle and pressure to produce the desired sound. While prior robotic cellists focus on accurate bowing trajectories, these works often rely on expensive motion capture techniques, and fail to sightread music in a human-like way. We propose a novel end-to-end MIDI score to robotic motion pipeline which converts musical input directly into collision-aware bowing motions for a UR5e robot cellist. Through use of Universal Robot Freedrive feature, our robotic musician can achieve human-like sound without the need for motion capture. Additionally, this work records live joint data via Real-Time Data Exchange (RTDE) as the robot plays, providing labeled robotic playing data from a collection of five standard pieces to the research community. To demonstrate the effectiveness of our method in comparison to human performers, we introduce the Musical Turing Test, in which a collection of 132 human participants evaluate our robot's performance against a human baseline. Human reference recordings are also released, enabling direct comparison for future studies. This evaluation technique establishes the first benchmark for robotic cello performance. Finally, we outline a residual reinforcement learning methodology to improve upon baseline robotic controls, highlighting future opportunities for improved string-crossing efficiency and sound quality.