Voices of Civilizations: A Multilingual QA Benchmark for Global Music Understanding

We introduce Voices of Civilizations, the first multilingual QA benchmark for evaluating audio LLMs' cultural comprehension on full-length music recordings. Covering 380 tracks across 38 languages, our automated pipeline yields 1,190 multiple-choice questions through four stages - each followed by manual verification: 1) compiling a representative music list; 2) generating cultural-background documents for each sample in the music list via LLMs; 3) extracting key attributes from those documents; and 4) constructing multiple-choice questions probing language, region associations, mood, and thematic content. We evaluate models under four conditions and report per-language accuracy. Our findings demonstrate that even state-of-the-art audio LLMs struggle to capture subtle cultural nuances without rich textual context and exhibit systematic biases in interpreting music from different cultural traditions. The dataset is publicly available on Hugging Face to foster culturally inclusive music understanding research.

DashengTokenizer: One layer is enough for unified audio understanding and generation

This paper introduces DashengTokenizer, a continuous audio tokenizer engineered for joint use in both understanding and generation tasks. Unlike conventional approaches, which train acoustic tokenizers and subsequently integrate frozen semantic knowledge, our method inverts this paradigm: we leverage frozen semantic features and inject acoustic information. In linear evaluation across 22 diverse tasks, our method outperforms previous audio codec and audio encoder baselines by a significant margin while maintaining competitive audio reconstruction quality. Notably, we demonstrate that this acoustic injection improves performance for tasks such as speech emotion recognition, music understanding, and acoustic scene classification. We further evaluate the tokenizer's generative performance on text-to-audio (TTA), text-to-music (TTM), and speech enhancement (SE). Our approach surpasses standard variational autoencoder (VAE)-based methods on TTA and TTM tasks, while its effectiveness on SE underscores its capabilities as a general-purpose audio encoder. Finally, our results challenge the prevailing assumption that VAE-based architectures are a prerequisite for audio synthesis. Checkpoints are available at https://huggingface.co/mispeech/dashengtokenizer.

Art2Mus: Artwork-to-Music Generation via Visual Conditioning and Large-Scale Cross-Modal Alignment

Music generation has advanced markedly through multimodal deep learning, enabling models to synthesize audio from text and, more recently, from images. However, existing image-conditioned systems suffer from two fundamental limitations: (i) they are typically trained on natural photographs, limiting their ability to capture the richer semantic, stylistic, and cultural content of artworks; and (ii) most rely on an image-to-text conversion stage, using language as a semantic shortcut that simplifies conditioning but prevents direct visual-to-audio learning. Motivated by these gaps, we introduce ArtSound, a large-scale multimodal dataset of 105,884 artwork-music pairs enriched with dual-modality captions, obtained by extending ArtGraph and the Free Music Archive. We further propose ArtToMus, the first framework explicitly designed for direct artwork-to-music generation, which maps digitized artworks to music without image-to-text translation or language-based semantic supervision. The framework projects visual embeddings into the conditioning space of a latent diffusion model, enabling music synthesis guided solely by visual information. Experimental results show that ArtToMus generates musically coherent and stylistically consistent outputs that reflect salient visual cues of the source artworks. While absolute alignment scores remain lower than those of text-conditioned systems-as expected given the substantially increased difficulty of removing linguistic supervision-ArtToMus achieves competitive perceptual quality and meaningful cross-modal correspondence. This work establishes direct visual-to-music generation as a distinct and challenging research direction, and provides resources that support applications in multimedia art, cultural heritage, and AI-assisted creative practice. Code and dataset will be publicly released upon acceptance.

Depth-Structured Music Recurrence: Budgeted Recurrent Attention for Full-Piece Symbolic Music Modeling

Long-context modeling is essential for symbolic music generation, since motif repetition and developmental variation can span thousands of musical events. However, practical composition and performance workflows frequently rely on resource-limited devices (e.g., electronic instruments and portable computers), making heavy memory and attention computation difficult to deploy. We introduce Depth-Structured Music Recurrence (DSMR), a recurrent long-context Transformer for full-piece symbolic music modeling that extends context beyond fixed-length excerpts via segment-level recurrence with detached cross-segment states, featuring a layer-wise memory-horizon schedule that budgets recurrent KV states across depth. DSMR is trained in a single left-to-right pass over each complete composition, akin to how a musician experiences it from beginning to end, while carrying recurrent cross-segment states forward. Within this recurrent framework, we systematically study how depth-wise horizon allocations affect optimization, best-checkpoint perplexity, and efficiency. By allocating different history-window lengths across layers while keeping the total recurrent-state budget fixed, DSMR creates depth-dependent temporal receptive fields within a recurrent attention stack without reducing compute depth. Our main instantiation is a two-scale DSMR schedule that allocates long history windows to lower layers and a uniform short window to the remaining layers. Experiments on the piano performance dataset MAESTRO demonstrate that two-scale DSMR provides a practical quality--efficiency recipe for full-length long-context symbolic music modeling with recurrent attention under limited computational resources.

Evaluating Disentangled Representations for Controllable Music Generation

Recent approaches in music generation rely on disentangled representations, often labeled as structure and timbre or local and global, to enable controllable synthesis. Yet the underlying properties of these embeddings remain underexplored. In this work, we evaluate such disentangled representations in a set of music audio models for controllable generation using a probing-based framework that goes beyond standard downstream tasks. The selected models reflect diverse unsupervised disentanglement strategies, including inductive biases, data augmentations, adversarial objectives, and staged training procedures. We further isolate specific strategies to analyze their effect. Our analysis spans four key axes: informativeness, equivariance, invariance, and disentanglement, which are assessed across datasets, tasks, and controlled transformations. Our findings reveal inconsistencies between intended and actual semantics of the embeddings, suggesting that current strategies fall short of producing truly disentangled representations, and prompting a re-examination of how controllability is approached in music generation.

SOM-VQ: Topology-Aware Tokenization for Interactive Generative Models

Vector-quantized representations enable powerful discrete generative models but lack semantic structure in token space, limiting interpretable human control. We introduce SOM-VQ, a tokenization method that combines vector quantization with Self-Organizing Maps to learn discrete codebooks with explicit low-dimensional topology. Unlike standard VQ-VAE, SOM-VQ uses topology-aware updates that preserve neighborhood structure: nearby tokens on a learned grid correspond to semantically similar states, enabling direct geometric manipulation of the latent space. We demonstrate that SOM-VQ produces more learnable token sequences in the evaluated domains while providing an explicit navigable geometry in code space. Critically, the topological organization enables intuitive human-in-the-loop control: users can steer generation by manipulating distances in token space, achieving semantic alignment without frame-level constraints. We focus on human motion generation - a domain where kinematic structure, smooth temporal continuity, and interactive use cases (choreography, rehabilitation, HCI) make topology-aware control especially natural - demonstrating controlled divergence and convergence from reference sequences through simple grid-based sampling. SOM-VQ provides a general framework for interpretable discrete representations applicable to music, gesture, and other interactive generative domains.

Stemphonic: All-at-once Flexible Multi-stem Music Generation

Music stem generation, the task of producing musically-synchronized and isolated instrument audio clips, offers the potential of greater user control and better alignment with musician workflows compared to conventional text-to-music models. Existing stem generation approaches, however, either rely on fixed architectures that output a predefined set of stems in parallel, or generate only one stem at a time, resulting in slow inference despite flexibility in stem combination. We propose Stemphonic, a diffusion-/flow-based framework that overcomes this trade-off and generates a variable set of synchronized stems in one inference pass. During training, we treat each stem as a batch element, group synchronized stems in a batch, and apply a shared noise latent to each group. At inference-time, we use a shared initial noise latent and stem-specific text inputs to generate synchronized multi-stem outputs in one pass. We further expand our approach to enable one-pass conditional multi-stem generation and stem-wise activity controls to empower users to iteratively generate and orchestrate the temporal layering of a mix. We benchmark our results on multiple open-source stem evaluation sets and show that Stemphonic produces higher-quality outputs while accelerating the full mix generation process by 25 to 50%. Demos at: https://stemphonic-demo.vercel.app.

DTT-BSR: GAN-based DTTNet with RoPE Transformer Enhancement for Music Source Restoration

Music source restoration (MSR) aims to recover unprocessed stems from mixed and mastered recordings. The challenge lies in both separating overlapping sources and reconstructing signals degraded by production effects such as compression and reverberation. We therefore propose DTT-BSR, a hybrid generative adversarial network (GAN) combining rotary positional embeddings (RoPE) transformer for long-term temporal modeling with dual-path band-split recurrent neural network (RNN) for multi-resolution spectral processing. Our model achieved 3rd place on the objective leaderboard and 4th place on the subjective leaderboard on the ICASSP 2026 MSR Challenge, demonstrating exceptional generation fidelity and semantic alignment with a compact size of 7.1M parameters.

Musical Metamerism with Time--Frequency Scattering

The concept of metamerism originates from colorimetry, where it describes a sensation of visual similarity between two colored lights despite significant differences in spectral content. Likewise, we propose to call ``musical metamerism'' the sensation of auditory similarity which is elicited by two music fragments which differ in terms of underlying waveforms. In this technical report, we describe a method to generate musical metamers from any audio recording. Our method is based on joint time--frequency scattering in Kymatio, an open-source software in Python which enables GPU computing and automatic differentiation. The advantage of our method is that it does not require any manual preprocessing, such as transcription, beat tracking, or source separation. We provide a mathematical description of JTFS as well as some excerpts from the Kymatio source code. Lastly, we review the prior work on JTFS and draw connections with closely related algorithms, such as spectrotemporal receptive fields (STRF), modulation power spectra (MPS), and Gabor filterbank (GBFB).

Video-based Music Generation

As the volume of video content on the internet grows rapidly, finding a suitable soundtrack remains a significant challenge. This thesis presents EMSYNC (EMotion and SYNChronization), a fast, free, and automatic solution that generates music tailored to the input video, enabling content creators to enhance their productions without composing or licensing music. Our model creates music that is emotionally and rhythmically synchronized with the video. A core component of EMSYNC is a novel video emotion classifier. By leveraging pretrained deep neural networks for feature extraction and keeping them frozen while training only fusion layers, we reduce computational complexity while improving accuracy. We show the generalization abilities of our method by obtaining state-of-the-art results on Ekman-6 and MovieNet. Another key contribution is a large-scale, emotion-labeled MIDI dataset for affective music generation. We then present an emotion-based MIDI generator, the first to condition on continuous emotional values rather than discrete categories, enabling nuanced music generation aligned with complex emotional content. To enhance temporal synchronization, we introduce a novel temporal boundary conditioning method, called "boundary offset encodings," aligning musical chords with scene changes. Combining video emotion classification, emotion-based music generation, and temporal boundary conditioning, EMSYNC emerges as a fully automatic video-based music generator. User studies show that it consistently outperforms existing methods in terms of music richness, emotional alignment, temporal synchronization, and overall preference, setting a new state-of-the-art in video-based music generation.