Search

Achieving extremely high-quality and truly immersive interactive Virtual Reality (VR) is expected to require a wireless link to the cloud, providing multi-gigabit throughput and extremely low latency. A prime candidate for fulfilling these requirements is millimeter-wave (mmWave) communications, operating in the 30 to 300 GHz bands, rather than the traditional sub-6 GHz. Evaluations with first-generation mmWave Wi-Fi hardware, based on the IEEE 802.11ad standard, have so far largely remained limited to lower-layer metrics. In this work, we present the first experimental analysis of the capabilities of mmWave for streaming VR content, using a novel testbed capable of repeatably creating blockage through mobility. Using this testbed, we show that (a) motion may briefly interrupt transmission, (b) a broken line of sight may degrade throughput unpredictably, and (c) TCP-based streaming frameworks need careful tuning to behave well over mmWave., Comment: Published and presented at International Conference on Network of the Future 2022

In modeling musical surprisal expectancy with computational methods, it has been proposed to use the information content (IC) of one-step predictions from an autoregressive model as a proxy for surprisal in symbolic music. With an appropriately chosen model, the IC of musical events has been shown to correlate with human perception of surprise and complexity aspects, including tonal and rhythmic complexity. This work investigates whether an analogous methodology can be applied to music audio. We train an autoregressive Transformer model to predict compressed latent audio representations of a pretrained autoencoder network. We verify learning effects by estimating the decrease in IC with repetitions. We investigate the mean IC of musical segment types (e.g., A or B) and find that segment types appearing later in a piece have a higher IC than earlier ones on average. We investigate the IC's relation to audio and musical features and find it correlated with timbral variations and loudness and, to a lesser extent, dissonance, rhythmic complexity, and onset density related to audio and musical features. Finally, we investigate if the IC can predict EEG responses to songs and thus model humans' surprisal in music. We provide code for our method on github.com/sonycslparis/audioic., Comment: 5 pages, 2 figures, 1 table. Accepted at the 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2025), Hyderabad, India

The freshness of sensor data is critical for all types of cyber-physical systems. An established measure for quantifying data freshness is the Age-of-Information (AoI), which has been the subject of extensive research. Recently, there has been increased interest in multi-sensor systems: redundant sensors producing samples of the same physical process, sensors such as cameras producing overlapping views, or distributed sensors producing correlated samples. When the information from a particular sensor is outdated, fresh samples from other correlated sensors can be helpful. To quantify the utility of distant but correlated samples, we put forth a two-dimensional (2D) model of AoI that takes into account the sensor distance in an age-equivalent representation. Since we define 2D-AoI as equivalent to AoI, it can be readily linked to existing AoI research, especially on parallel systems. We consider physical phenomena modeled as spatio-temporal processes and derive the 2D-AoI for different Gaussian correlation kernels. For a basic exponential product kernel, we find that spatial distance causes an additive offset of the AoI, while for other kernels the effects of spatial distance are more complex and vary with time. Using our methodology, we evaluate the 2D-AoI of different spatial topologies and sensor densities.

Treatment of cancer has rapidly evolved over time in quite dramatic ways, for example from chemotherapies, targeted therapies to immunotherapies and chimeric antigen receptor T-cells. Nonetheless, the basic design of early phase I trials in oncology still follows pre-dominantly a dose-escalation design. These trials monitor safety over the first treatment cycle in order to escalate the dose of the investigated drug. However, over time studying additional factors such as drug combinations and/or variation in the timing of dosing became important as well. Existing designs were continuously enhanced and expanded to account for increased trial complexity. With toxicities occurring at later stages beyond the first cycle and the need to treat patients over multiple cycles, the focus on the first treatment cycle only is becoming a limitation in nowadays multi-cycle treatment therapies. Here we introduce a multi-cycle time-to-event model (TITE-CLRM: Time-Interval-To-Event Complementary-Loglog Regression Model) allowing guidance of dose-escalation trials studying multi-cycle therapies. The challenge lies in balancing the need to monitor safety of longer treatment periods with the need to continuously enroll patients safely. The proposed multi-cycle time to event model is formulated as an extension to established concepts like the escalation with over dose control principle. The model is motivated from a current drug development project and evaluated in a simulation study., Comment: 18 pages, 8 figures, 1 table, 4 supplementary figures

Music therapy has been shown in recent years to provide multiple health benefits related to emotional wellness. In turn, maintaining a healthy emotional state has proven to be effective for patients undergoing treatment, such as Parkinson's patients or patients suffering from stress and anxiety. We propose fine-tuning MusicGen, a music-generating transformer model, to create short musical clips that assist patients in transitioning from negative to desired emotional states. Using low-rank decomposition fine-tuning on the MTG-Jamendo Dataset with emotion tags, we generate 30-second clips that adhere to the iso principle, guiding patients through intermediate states in the valence-arousal circumplex. The generated music is evaluated using a music emotion recognition model to ensure alignment with intended emotions. By concatenating these clips, we produce a 15-minute "music medicine" resembling a music therapy session. Our approach is the first model to leverage Language Models to generate music medicine. Ultimately, the output is intended to be used as a temporary relief between music therapy sessions with a board-certified therapist., Comment: Late-Breaking / Demo Session Extended Abstract, ISMIR 2024 Conference

We investigate tearing modes (TM) driven by current density gradient in collisionless tokamak plasmas by using the electromagnetic gyrokinetic simulation code ORB5. We elucidate the TM width by simulations for flat profiles, as the absence of background diamagnetic flows implies a small rotation-speed, while finite-gradients are included to investigate the TM rotation. For flat profiles, the initial saturation width of nonlinearly driven magnetic islands is related to the TM linear growth rate; however, large islands in the initial saturation phase are prone to current density redistribution that reduces the island width in the following evolution. Island-induced $\boldsymbol{E}\times\boldsymbol{B}$ and diamagnetic sheared flows develop at the separatrix, able to destabilize the Kelvin-Helmholtz instability (KHI). The KHI turbulence enhances a strong quadrupole vortex flow that reinforces the island decay, resulting in a strong reduction of the island width in an eventual steady state. This process is enhanced by trapped electrons. For finite gradients profile, the TM usually rotates in the electron diamagnetic direction, but can change direction when the ion temperature gradient dominates the other gradients. The reduced growth of the TM by diamagnetic effects results in a moderate island size, which remains almost unchanged after the initial saturation. At steady state, strong zonal flows are nonlinearly excited and dominate the island rotation, as expected from previous theoretical and numerical studies. When the plasma beta is increased, the TM mode is suppressed and a mode with the same helicity but with twisting parity, coupled with the neighboring poloidal harmonics, is destabilized, similar to the kinetic ballooning mode.

Recent advances in nanomaterials have pushed the boundaries of nanoscale fabrication to the limit of single atoms (SAs), particularly in heterogeneous catalysis. Single atom catalysts (SACs), comprising minute amounts of transition metals dispersed on inert substrates, have emerged as prominent materials in this domain. However, overcoming the tendency of these SAs to cluster beyond cryogenic temperatures and precisely arranging them on surfaces pose significant challenges. Employing organic templates for orchestrating and modulating the activity of single atoms holds promise. Here, we introduce a novel single atom platform (SAP) wherein atoms are firmly anchored to specific coordination sites distributed along carbon-based polymers, synthesized via on-surface synthesis (OSS). These SAPs exhibit atomiclevel structural precision and stability, even at elevated temperatures. The asymmetry in the electronic states at the active sites anticipates the enhanced reactivity of these precisely defined reactive centers. Upon exposure to CO and CO2 gases at low temperatures, the SAP demonstrates excellent trapping capabilities. Fine-tuning the structure and properties of the coordination sites offers unparalleled flexibility in tailoring functionalities, thus opening avenues for previously untapped potential in catalytic applications., Comment: Main text (12 pages, 4 figures) and supplementary information (15 pages, 16 figures)

We show that, in the space of all totally real fields equipped with the constructible topology, the set of fields that admit a universal quadratic form, or have the Northcott property, is meager. The main tool is a new theorem on the number of square classes of totally positive units represented by a quadratic lattice of a given rank., Comment: preprint, 12 pages

We propose a pre-training pipeline for audio spectrogram transformers for frame-level sound event detection tasks. On top of common pre-training steps, we add a meticulously designed training routine on AudioSet frame-level annotations. This includes a balanced sampler, aggressive data augmentation, and ensemble knowledge distillation. For five transformers, we obtain a substantial performance improvement over previously available checkpoints both on AudioSet frame-level predictions and on frame-level sound event detection downstream tasks, confirming our pipeline's effectiveness. We publish the resulting checkpoints that researchers can directly fine-tune to build high-performance models for sound event detection tasks., Comment: Submitted to ICASSP'25. Source code available: https://github.com/fschmid56/PretrainedSED

We characterize the self-assembly and phase behavior of Janus rods over a broad range of temperatures and volume fractions, using Langevin dynamics simulations and free energy calculations. The Janus rods consist of a line of fused overlapping spheres that interact via a soft-core repulsive potential, with the addition of an attractive pseudo-square-well tail to a fraction of the spheres (the coverage) ranging from 5\% to 100\% of sites. Competition between the stability of liquid crystal phases originating from shape anisotropy and assembly driven by directional interactions gives rise to a rich polymorphism that depends on the coverage. At low density near the Boyle temperature, we observe the formation of spherical and tubular micelles at low coverage, while at higher coverage randomly oriented monolayers form as the attractive parts of the rods overlap. At higher density, bilayer structures appear and merge to form smectic and crystalline lamellar phases. All of these structures gradually become unstable as the temperature is increased until eventually regular nematic and smectic phases appear, consistent with the hard rod limit. Our results indicate that the intermediate regime where shape-entropic effects compete with anisotropic attractions provided by site specificity is rich in structural possibilities, and should help guide the design of rod-like colloids for specific applications., Comment: 13 pages, 13 figures, SI not included

Dual-encoder-based audio retrieval systems are commonly optimized with contrastive learning on a set of matching and mismatching audio-caption pairs. This leads to a shared embedding space in which corresponding items from the two modalities end up close together. Since audio-caption datasets typically only contain matching pairs of recordings and descriptions, it has become common practice to create mismatching pairs by pairing the audio with a caption randomly drawn from the dataset. This is not ideal because the randomly sampled caption could, just by chance, partly or entirely describe the audio recording. However, correspondence information for all possible pairs is costly to annotate and thus typically unavailable; we, therefore, suggest substituting it with estimated correspondences. To this end, we propose a two-staged training procedure in which multiple retrieval models are first trained as usual, i.e., without estimated correspondences. In the second stage, the audio-caption correspondences predicted by these models then serve as prediction targets. We evaluate our method on the ClothoV2 and the AudioCaps benchmark and show that it improves retrieval performance, even in a restricting self-distillation setting where a single model generates and then learns from the estimated correspondences. We further show that our method outperforms the current state of the art by 1.6 pp. mAP@10 on the ClothoV2 benchmark., Comment: In Proceedings of the 9th Workshop on Detection and Classification of Acoustic Scenes and Events, DCASE, Tokyo, Japan, 2024. Implementation available on GitHub: https://github.com/OptimusPrimus/salsa

Query-by-Vocal Imitation (QBV) is about searching audio files within databases using vocal imitations created by the user's voice. Since most humans can effectively communicate sound concepts through voice, QBV offers the more intuitive and convenient approach compared to text-based search. To fully leverage QBV, developing robust audio feature representations for both the vocal imitation and the original sound is crucial. In this paper, we present a new system for QBV that utilizes the feature extraction capabilities of Convolutional Neural Networks pre-trained with large-scale general-purpose audio datasets. We integrate these pre-trained models into a dual encoder architecture and fine-tune them end-to-end using contrastive learning. A distinctive aspect of our proposed method is the fine-tuning strategy of pre-trained models using an adapted NT-Xent loss for contrastive learning, creating a shared embedding space for reference recordings and vocal imitations. The proposed system significantly enhances audio retrieval performance, establishing a new state of the art on both coarse- and fine-grained QBV tasks., Comment: Accepted to the DCASE Workshop 2024. Source code available: https://github.com/Jonathan-Greif/QBV

In recent years, the quality and public interest in music generation systems have grown, encouraging research into various ways to control these systems. We propose a novel method for controlling surprisal in music generation using sequence models. To achieve this goal, we define a metric called Instantaneous Information Content (IIC). The IIC serves as a proxy function for the perceived musical surprisal (as estimated from a probabilistic model) and can be calculated at any point within a music piece. This enables the comparison of surprisal across different musical content even if the musical events occur in irregular time intervals. We use beam search to generate musical material whose IIC curve closely approximates a given target IIC. We experimentally show that the IIC correlates with harmonic and rhythmic complexity and note density. The correlation decreases with the length of the musical context used for estimating the IIC. Finally, we conduct a qualitative user study to test if human listeners can identify the IIC curves that have been used as targets when generating the respective musical material. We provide code for creating IIC interpolations and IIC visualizations on https://github.com/muthissar/iic., Comment: 8 pages, 4 figures, 2 tables, accepted at the 25th Int. Society for Music Information Retrieval Conf., San Francisco, USA, 2024

Automatic Music Transcription (AMT) is the task of recognizing notes in audio recordings of music. The State-of-the-Art (SotA) benchmarks have been dominated by deep learning systems. Due to the scarcity of high quality data, they are usually trained and evaluated exclusively or predominantly on classical piano music. Unfortunately, that hinders our ability to understand how they generalize to other music. Previous works have revealed several aspects of memorization and overfitting in these systems. We identify two primary sources of distribution shift: the music, and the sound. Complementing recent results on the sound axis (i.e. acoustics, timbre), we investigate the musical one (i.e. note combinations, dynamics, genre). We evaluate the performance of several SotA AMT systems on two new experimental test sets which we carefully construct to emulate different levels of musical distribution shift. Our results reveal a stark performance gap, shedding further light on the Corpus Bias problem, and the extent to which it continues to trouble these systems., Comment: 2 pages, 1 figure, presented in the 1st International Workshop on Sound Signal Processing Applications (IWSSPA) 2024

Note alignment refers to the task of matching individual notes of two versions of the same symbolically encoded piece. Methods addressing this task commonly rely on sequence alignment algorithms such as Hidden Markov Models or Dynamic Time Warping (DTW) applied directly to note or onset sequences. While successful in many cases, such methods struggle with large mismatches between the versions. In this work, we learn note-wise representations from data augmented with various complex mismatch cases, e.g. repeats, skips, block insertions, and long trills. At the heart of our approach lies a transformer encoder network - TheGlueNote - which predicts pairwise note similarities for two 512 note subsequences. We postprocess the predicted similarities using flavors of weightedDTW and pitch-separated onsetDTW to retrieve note matches for two sequences of arbitrary length. Our approach performs on par with the state of the art in terms of note alignment accuracy, is considerably more robust to version mismatches, and works directly on any pair of MIDI files., Comment: to be published in Proceedings of the 25th International Society for Music Information Retrieval Conference (ISMIR), 2024

The power of distributed acoustic sensing (DAS) lies in its ability to sample deformation signals along an optical fiber at hundreds of locations with only one interrogation unit (IU). While the IU is calibrated to record 'fiber strain', the properties of the cable and its coupling to the rock control the 'strain transfer rate' and hence how much of 'rock strain' is represented in the recorded signal. We use DAS recordings in an underground installation near an array of strainmeters in order to calibrate the 'strain transfer rate' in situ, using earthquake signals between 0.05 Hz and 0.1 Hz. A tight-buffered cable and a standard loose-tube telecommunication cable (running in parallel) are used, where a section of both cables loaded down by loose sand and sand bags is compared to a section, where cables are just unreeled on the floor. The 'strain transfer rate' varies between 0.13 and 0.53 depending on cable and installation type. The sandbags show no obvious effect and the tight-buffered cable generally provides a larger 'strain transfer rate'. Calibration of the 'strain transfer rate' with respect to the strainmeter does not depend on wave propagation parameters. Hence it is applicable to the large amplitude surface wave signal in a strain component almost perpendicular to the great-circle direction for which a waveform comparison with seismometer data does not work. The noise background for 'rock strain' in the investigated band is found at about an rms-amplitude of 0.1 nstrain in 1/6 decade for the tight-buffered cable. This allows a detection of marine microseisms at times of high microseism amplitude.

We propose a system for tracking beats and downbeats with two objectives: generality across a diverse music range, and high accuracy. We achieve generality by training on multiple datasets -- including solo instrument recordings, pieces with time signature changes, and classical music with high tempo variations -- and by removing the commonly used Dynamic Bayesian Network (DBN) postprocessing, which introduces constraints on the meter and tempo. For high accuracy, among other improvements, we develop a loss function tolerant to small time shifts of annotations, and an architecture alternating convolutions with transformers either over frequency or time. Our system surpasses the current state of the art in F1 score despite using no DBN. However, it can still fail, especially for difficult and underrepresented genres, and performs worse on continuity metrics, so we publish our model, code, and preprocessed datasets, and invite others to beat this., Comment: Accepted at the 25th International Society for Music Information Retrieval Conference (ISMIR), 2024

This technical report describes the CP-JKU team's submission for Task 4 Sound Event Detection with Heterogeneous Training Datasets and Potentially Missing Labels of the DCASE 24 Challenge. We fine-tune three large Audio Spectrogram Transformers, PaSST, BEATs, and ATST, on the joint DESED and MAESTRO datasets in a two-stage training procedure. The first stage closely matches the baseline system setup and trains a CRNN model while keeping the large pre-trained transformer model frozen. In the second stage, both CRNN and transformer are fine-tuned using heavily weighted self-supervised losses. After the second stage, we compute strong pseudo-labels for all audio clips in the training set using an ensemble of all three fine-tuned transformers. Then, in a second iteration, we repeat the two-stage training process and include a distillation loss based on the pseudo-labels, boosting single-model performance substantially. Additionally, we pre-train PaSST and ATST on the subset of AudioSet that comes with strong temporal labels, before fine-tuning them on the Task 4 datasets., Comment: Technical Report describing our system for DCASE2024 Challenge Task 4: https://dcase.community/challenge2024/task-sound-event-detection-with-heterogeneous-training-dataset-and-potentially-missing-labels-results Code: https://github.com/CPJKU/cpjku_dcase24. arXiv admin note: text overlap with arXiv:2407.12997

A central problem in building effective sound event detection systems is the lack of high-quality, strongly annotated sound event datasets. For this reason, Task 4 of the DCASE 2024 challenge proposes learning from two heterogeneous datasets, including audio clips labeled with varying annotation granularity and with different sets of possible events. We propose a multi-iteration, multi-stage procedure for fine-tuning Audio Spectrogram Transformers on the joint DESED and MAESTRO Real datasets. The first stage closely matches the baseline system setup and trains a CRNN model while keeping the pre-trained transformer model frozen. In the second stage, both CRNN and transformer are fine-tuned using heavily weighted self-supervised losses. After the second stage, we compute strong pseudo-labels for all audio clips in the training set using an ensemble of fine-tuned transformers. Then, in a second iteration, we repeat the two-stage training process and include a distillation loss based on the pseudo-labels, achieving a new single-model, state-of-the-art performance on the public evaluation set of DESED with a PSDS1 of 0.692. A single model and an ensemble, both based on our proposed training procedure, ranked first in Task 4 of the DCASE Challenge 2024., Comment: Code: https://github.com/CPJKU/cpjku_dcase24

Graph Neural Networks (GNNs) have recently gained traction in symbolic music tasks, yet a lack of a unified framework impedes progress. Addressing this gap, we present GraphMuse, a graph processing framework and library that facilitates efficient music graph processing and GNN training for symbolic music tasks. Central to our contribution is a new neighbor sampling technique specifically targeted toward meaningful behavior in musical scores. Additionally, GraphMuse integrates hierarchical modeling elements that augment the expressivity and capabilities of graph networks for musical tasks. Experiments with two specific musical prediction tasks -- pitch spelling and cadence detection -- demonstrate significant performance improvement over previous methods. Our hope is that GraphMuse will lead to a boost in, and standardization of, symbolic music processing based on graph representations. The library is available at https://github.com/manoskary/graphmuse, Comment: Accepted at the 25th International Society for Music Information Retrieval Conference (ISMIR 2024)

This paper approaches the problem of separating the notes from a quantized symbolic music piece (e.g., a MIDI file) into multiple voices and staves. This is a fundamental part of the larger task of music score engraving (or score typesetting), which aims to produce readable musical scores for human performers. We focus on piano music and support homophonic voices, i.e., voices that can contain chords, and cross-staff voices, which are notably difficult tasks that have often been overlooked in previous research. We propose an end-to-end system based on graph neural networks that clusters notes that belong to the same chord and connects them with edges if they are part of a voice. Our results show clear and consistent improvements over a previous approach on two datasets of different styles. To aid the qualitative analysis of our results, we support the export in symbolic music formats and provide a direct visualization of our outputs graph over the musical score. All code and pre-trained models are available at https://github.com/CPJKU/piano_svsep, Comment: Accepted at the 25th International Society for Music Information Retrieval (ISMIR) 2024

On-device inference holds great potential for increased energy efficiency, responsiveness, and privacy in edge ML systems. However, due to less capable ML models that can be embedded in resource-limited devices, use cases are limited to simple inference tasks such as visual keyword spotting, gesture recognition, and predictive analytics. In this context, the Hierarchical Inference (HI) system has emerged as a promising solution that augments the capabilities of the local ML by offloading selected samples to an edge server or cloud for remote ML inference. Existing works demonstrate through simulation that HI improves accuracy. However, they do not account for the latency and energy consumption on the device, nor do they consider three key heterogeneous dimensions that characterize ML systems: hardware, network connectivity, and models. In contrast, this paper systematically compares the performance of HI with on-device inference based on measurements of accuracy, latency, and energy for running embedded ML models on five devices with different capabilities and three image classification datasets. For a given accuracy requirement, the HI systems we designed achieved up to 73% lower latency and up to 77% lower device energy consumption than an on-device inference system. The key to building an efficient HI system is the availability of small-size, reasonably accurate on-device models whose outputs can be effectively differentiated for samples that require remote inference. Despite the performance gains, HI requires on-device inference for all samples, which adds a fixed overhead to its latency and energy consumption. Therefore, we design a hybrid system, Early Exit with HI (EE-HI), and demonstrate that compared to HI, EE-HI reduces the latency by up to 59.7% and lowers the device's energy consumption by up to 60.4%.

Integrated Sensing And Communication (ISAC ) systems are expected to perform accurate radar sensing while having minimal impact on communication. Ideally, sensing should only reuse communication resources, especially for spectrum which is contended by many applications. However, this poses a great challenge in that communication systems often operate on narrow subbands with low sensing resolution. Combining contiguous subbands has shown significant resolution gain in active localization. However, multiband ISAC remains unexplored due to communication subbands being highly sparse (non-contiguous) and affected by phase offsets that prevent their aggregation (incoherent). To tackle these problems, we design HiSAC, the first multiband ISAC system that combines diverse subbands across a wide frequency range to achieve super-resolved passive ranging. To solve the non-contiguity and incoherence of subbands, HiSAC combines them progressively, exploiting an anchor propagation path between transmitter and receiver in an optimization problem to achieve phase coherence. HiSAC fully reuses pilot signals in communication systems, it applies to different frequencies and can combine diverse technologies, e.g., 5G-NR and WiGig. We implement HiSAC on an experimental platform in the millimeter-wave unlicensed band and test it on objects and humans. Our results show it enhances the sensing resolution by up to 20 times compared to single-band processing while occupying the same spectrum., Comment: 15 pages, 15 figures

Matching raw audio signals with textual descriptions requires understanding the audio's content and the description's semantics and then drawing connections between the two modalities. This paper investigates a hybrid retrieval system that utilizes audio metadata as an additional clue to understand the content of audio signals before matching them with textual queries. We experimented with metadata often attached to audio recordings, such as keywords and natural-language descriptions, and we investigated late and mid-level fusion strategies to merge audio and metadata. Our hybrid approach with keyword metadata and late fusion improved the retrieval performance over a content-based baseline by 2.36 and 3.69 pp. mAP@10 on the ClothoV2 and AudioCaps benchmarks, respectively., Comment: In Proceedings of the 32nd European Signal Processing Conference, EUSIPCO 2024

In the pursuit of developing expressive music performance models using artificial intelligence, this paper introduces DExter, a new approach leveraging diffusion probabilistic models to render Western classical piano performances. In this approach, performance parameters are represented in a continuous expression space and a diffusion model is trained to predict these continuous parameters while being conditioned on the musical score. Furthermore, DExter also enables the generation of interpretations (expressive variations of a performance) guided by perceptually meaningful features by conditioning jointly on score and perceptual feature representations. Consequently, we find that our model is useful for learning expressive performance, generating perceptually steered performances, and transferring performance styles. We assess the model through quantitative and qualitative analyses, focusing on specific performance metrics regarding dimensions like asynchrony and articulation, as well as through listening tests comparing generated performances with different human interpretations. Results show that DExter is able to capture the time-varying correlation of the expressive parameters, and compares well to existing rendering models in subjectively evaluated ratings. The perceptual-feature-conditioned generation and transferring capabilities of DExter are verified by a proxy model predicting perceptual characteristics of differently steered performances., Comment: in submission to appsci special session

Large language models such as ChatGPT exhibit striking political biases. If users query them about political information, they often take a normative stance. To overcome this, we align LLMs with diverse political viewpoints from 100,000 comments written by candidates running for national parliament in Switzerland. Models aligned with this data can generate more accurate political viewpoints from Swiss parties, compared to commercial models such as ChatGPT. We also propose a procedure to generate balanced overviews summarizing multiple viewpoints using such models. The replication package contains all code and data., Comment: accepted at EMNLP 2024 main as a short paper

Automatic piano transcription models are typically evaluated using simple frame- or note-wise information retrieval (IR) metrics. Such benchmark metrics do not provide insights into the transcription quality of specific musical aspects such as articulation, dynamics, or rhythmic precision of the output, which are essential in the context of expressive performance analysis. Furthermore, in recent years, MAESTRO has become the de-facto training and evaluation dataset for such models. However, inference performance has been observed to deteriorate substantially when applied on out-of-distribution data, thereby questioning the suitability and reliability of transcribed outputs from such models for specific MIR tasks. In this work, we investigate the performance of three state-of-the-art piano transcription models in two experiments. In the first one, we propose a variety of musically informed evaluation metrics which, in contrast to the IR metrics, offer more detailed insight into the musical quality of the transcriptions. In the second experiment, we compare inference performance on real-world and perturbed audio recordings, and highlight musical dimensions which our metrics can help explain. Our experimental results highlight the weaknesses of existing piano transcription metrics and contribute to a more musically sound error analysis of transcription outputs., Comment: Accepted at the 25th International Society for Music Information Retrieval Conference (ISMIR 2024)

The current standard model of cosmology successfully describes a variety of measurements, but the nature of its main ingredients, dark matter and dark energy, remains unknown. Euclid is a medium-class mission in the Cosmic Vision 2015-2025 programme of the European Space Agency (ESA) that will provide high-resolution optical imaging, as well as near-infrared imaging and spectroscopy, over about 14,000 deg^2 of extragalactic sky. In addition to accurate weak lensing and clustering measurements that probe structure formation over half of the age of the Universe, its primary probes for cosmology, these exquisite data will enable a wide range of science. This paper provides a high-level overview of the mission, summarising the survey characteristics, the various data-processing steps, and data products. We also highlight the main science objectives and expected performance., Comment: Accepted for publication in the A&A special issue`Euclid on Sky'

This article describes the Data-Efficient Low-Complexity Acoustic Scene Classification Task in the DCASE 2024 Challenge and the corresponding baseline system. The task setup is a continuation of previous editions (2022 and 2023), which focused on recording device mismatches and low-complexity constraints. This year's edition introduces an additional real-world problem: participants must develop data-efficient systems for five scenarios, which progressively limit the available training data. The provided baseline system is based on an efficient, factorized CNN architecture constructed from inverted residual blocks and uses Freq-MixStyle to tackle the device mismatch problem. The task received 37 submissions from 17 teams, with the large majority of systems outperforming the baseline. The top-ranked system's accuracy ranges from 54.3% on the smallest to 61.8% on the largest subset, corresponding to relative improvements of approximately 23% and 9% over the baseline system on the evaluation set., Comment: Task Description Page: https://dcase.community/challenge2024/task-data-efficient-low-complexity-acoustic-scene-classification

In this work, we present Score MUsic Graph (SMUG)-Explain, a framework for generating and visualizing explanations of graph neural networks applied to arbitrary prediction tasks on musical scores. Our system allows the user to visualize the contribution of input notes (and note features) to the network output, directly in the context of the musical score. We provide an interactive interface based on the music notation engraving library Verovio. We showcase the usage of SMUG-Explain on the task of cadence detection in classical music. All code is available on https://github.com/manoskary/SMUG-Explain., Comment: In Proceedings of the Sound and Music Computing Conference 2024 (SMC2024), Porto, Portugal

We propose a new graph convolutional block, called MusGConv, specifically designed for the efficient processing of musical score data and motivated by general perceptual principles. It focuses on two fundamental dimensions of music, pitch and rhythm, and considers both relative and absolute representations of these components. We evaluate our approach on four different musical understanding problems: monophonic voice separation, harmonic analysis, cadence detection, and composer identification which, in abstract terms, translate to different graph learning problems, namely, node classification, link prediction, and graph classification. Our experiments demonstrate that MusGConv improves the performance on three of the aforementioned tasks while being conceptually very simple and efficient. We interpret this as evidence that it is beneficial to include perception-informed processing of fundamental musical concepts when developing graph network applications on musical score data., Comment: Accepted at the 33rd International Joint Conference on Artificial Intelligence (IJCAI-24)

Implantable Brain-machine interfaces (BMIs) are promising for motor rehabilitation and mobility augmentation, and they demand accurate and energy-efficient algorithms. In this paper, we propose a novel spiking neural network (SNN) decoder for regression tasks for implantable BMIs. The SNN is trained with enhanced spatio-temporal backpropagation to fully leverage its capability to handle temporal problems. The proposed SNN decoder outperforms the state-of-the-art Kalman filter and artificial neural network (ANN) decoders in offline finger velocity decoding tasks. The decoder is deployed on a RISC-V-based hardware platform and optimized to exploit sparsity. The proposed implementation has an average power consumption of 0.50 mW in a duty-cycled mode. When conducting continuous inference without duty-cycling, it achieves an energy efficiency of 1.88 uJ per inference, which is 5.5X less than the baseline ANN. Additionally, the average decoding latency is 0.12 ms for each inference, which is 5.7X faster than the ANN implementation.

Atomically precise graphene nanoribbons (GNRs) have a wide range of electronic properties that depend sensitively on their chemical structure. Several types of GNRs have been synthesized on metal surfaces through selective surface-catalyzed reactions. The resulting GNRs are adsorbed on the metal surface, which may lead to hybridization between the GNR orbitals and those of the substrate. This makes investigation of the intrinsic electronic properties of GNRs more difficult, and also rules out capacitive gating. Here we demonstrate the formation of a dielectric gold chloride adlayer that can intercalate underneath GNRs on the Au(111) surface. The intercalated gold chloride adlayer electronically decouples the GNRs from the metal and leads to a substantial hole doping of the GNRs. Our results introduce an easily accessible tool in the in situ characterization of GNRs grown on Au(111) that allows for exploration of their electronic properties in a heavily hole-doped regime., Comment: Supporting information follows the main text in the same file

Explainable Artificial Intelligence (XAI) poses a significant challenge in providing transparent and understandable insights into complex AI models. Traditional post-hoc algorithms, while useful, often struggle to deliver interpretable explanations. Concept-based models offer a promising avenue by incorporating explicit representations of concepts to enhance interpretability. However, existing research on automatic concept discovery methods is often limited by lower-level concepts, costly human annotation requirements, and a restricted domain of background knowledge. In this study, we explore the potential of a Large Language Model (LLM), specifically GPT-4, by leveraging its domain knowledge and common-sense capability to generate high-level concepts that are meaningful as explanations for humans, for a specific setting of image classification. We use minimal textual object information available in the data via prompting to facilitate this process. To evaluate the output, we compare the concepts generated by the LLM with two other methods: concepts generated by humans and the ECII heuristic concept induction system. Since there is no established metric to determine the human understandability of concepts, we conducted a human study to assess the effectiveness of the LLM-generated concepts. Our findings indicate that while human-generated explanations remain superior, concepts derived from GPT-4 are more comprehensible to humans compared to those generated by ECII.

The Hierarchical Inference (HI) paradigm employs a tiered processing: the inference from simple data samples are accepted at the end device, while complex data samples are offloaded to the central servers. HI has recently emerged as an effective method for balancing inference accuracy, data processing, transmission throughput, and offloading cost. This approach proves particularly efficient in scenarios involving resource-constrained edge devices, such as IoT sensors and micro controller units (MCUs), tasked with executing tinyML inference. Notably, it outperforms strategies such as local inference execution, inference offloading to edge servers or cloud facilities, and split inference (i.e., inference execution distributed between two endpoints). Building upon the HI paradigm, this work explores different techniques aimed at further optimizing inference task execution. We propose and discuss three distinct HI approaches and evaluate their utility for image classification.

mmWave communication has come up as the unexplored spectrum for 5G services. With new standards for 5G NR positioning, more off-the-shelf platforms and algorithms are needed to perform indoor positioning. An object can be accurately positioned in a room either by using an angle and a delay estimate or two angle estimates or three delay estimates. We propose an algorithm to jointly estimate the angle of arrival (AoA) and angle of departure (AoD), based only on the received signal strength (RSS). We use mm-FLEX, an experimentation platform developed by IMDEA Networks Institute that can perform real-time signal processing for experimental validation of our proposed algorithm. Codebook-based beampatterns are used with a uniquely placed multi-antenna array setup to enhance the reception of multipath components and we obtain an AoA estimate per receiver thereby overcoming the line-of-sight (LoS) limitation of RSS-based localization systems. We further validate the results from measurements by emulating the setup with a simple ray-tracing approach., Comment: Accepted in IEEE WCNC 2024

In this work, we compare existing non-polarizable force fields developed to study the solid or solution phases of hybrid organic-inorganic halide perovskites with the AMOEBA polarizable force field. The aim is to test whether more computationally expensive polarizable force fields like AMOEBA offer better transferability between solution and solid phases, with the ultimate goal being the study of crystal nucleation, growth and other interfacial phenomena involving these ionic compounds. In the context of hybrid perovskites, AMOEBA force field parameters already exist for several elements in solution and we decided to leave them unchanged and to only parameterize the missing ones (Pb\textsuperscript{2+} and CH\textsubscript{3}NH\textsubscript{3}\textsuperscript{+} ions) in order to maximise transferability and avoid over-fitting to the specific examples studied here. Overall, we find that AMOEBA yields accurate hydration free energies (within 5\%) for typical ionic species while showing the correct ordering of stability for the different crystal polymorphs of CsPbI\textsubscript{3} and CH\textsubscript{3}NH\textsubscript{3}PbI\textsubscript{3}. While the existing parameters do not accurately reproduce all transition temperatures and lattice parameters, AMOEBA offers better transferability between solution and solid states than existing non-polarizable force fields.

The electrification of public transport vehicles offers the potential to relieve city centers of pollutant and noise emissions. Furthermore, electric buses have lower life-cycle greenhouse gas (GHG) emissions than diesel buses, particularly when operated with sustainably produced electricity. However, the heating, ventilation, and air-conditioning (HVAC) system can consume a significant amount of energy, thus limiting the achievable driving range. In this paper, we address the HVAC system in an electric city bus by analyzing the trade-off between the energy consumption and the thermal comfort of the passengers. We do this by developing a dynamic thermal model for the bus, which we simplify by considering it to be in steady state. We introduce a method that is able to quickly optimize the steady-state HVAC system inputs for a large number of samples representative of a year-round operation. A comparison between the results from the steady-state optimization approach and a dynamic simulation reveals small deviations in both the HVAC system power demand and achieved thermal comfort. Thus, the approximation of the system performance with a steady-state model is justified. We present two case studies to demonstrate the practical relevance of the approach. First, we show how the method can be used to compare different HVAC system designs based on a year-round performance evaluation. Second, we show how the method can be used to extract setpoints for online controllers that achieve close-to-optimal performance without any predictive information. In conclusion, this study shows that a steady-state analysis of the HVAC systems of an electric city bus is a valuable approach to evaluate and optimize its performance., Comment: Preprint submitted to Control Engineering Practice