Tootfinder

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
- Graph Neural Network for Neutrino Physics Event Reconstruction
Hewes, Aurisano, Cerati, Kowalkowski, Lee, Liao, Grzenda, Gumpula, Zhang

Video-based Direct Time Series Measurement of Along-Strike Slip on the Coseismic Surface Rupture During the 2025 Mw7.7 Myanmar Earthquake
Jianhao Gao, Fuhua Zheng, Chaofeng Wang, Haoran Meng
https://arxiv.org/abs/2505.20494

Video-based Direct Time Series Measurement of Along-Strike Slip on the Coseismic Surface Rupture During the 2025 Mw7.7 Myanmar Earthquake
This study presents a time-resolved analysis of coseismic lateral surface rupture along the Sagaing Fault during the Mw 7.7 Mandalay, Myanmar earthquake on March 28, 2025. Leveraging a publicly available Closed-Circuit Television (CCTV) footage alongside on-site measurements, we show the first in-situ high sampling rate direct measurement of a coseismic slip evolution of a fault during an earthquake. Our work comprises four primary stages: data acquisition, video pre-processing, object tracking…

The world's data centers already consume three times as much electricity as Dutch national consumption. And this is expected to grow to 8 times within five years.
https://mastodon.green/@solar_chase/114556281180000861

Eventual, which develops Daft, a Python-native open-source data processing engine, raised a $7.5M seed led by CRV and a $20M Series A led by Felicis (Rebecca Szkutak/TechCrunch)
https://techcrunch.com/2025/06/24/how-a-data-pro…

How a data processing problem at Lyft became the basis for Eventual | TechCrunch
Eventual's data processing engine Daft was inspried by the founders' experience working on Lyft's autonomous vehicle project.

This https://arxiv.org/abs/2505.03071 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

The Search for Squawk: Agile Modeling in Bioacoustics
Passive acoustic monitoring (PAM) has shown great promise in helping ecologists understand the health of animal populations and ecosystems. However, extracting insights from millions of hours of audio recordings requires the development of specialized recognizers. This is typically a challenging task, necessitating large amounts of training data and machine learning expertise. In this work, we introduce a general, scalable and data-efficient system for developing recognizers for novel bioacoust…

This https://arxiv.org/abs/2502.03210 has been replaced.
initial toot: https://mastoxiv.page/@arX…

From Kernels to Features: A Multi-Scale Adaptive Theory of Feature Learning
Feature learning in neural networks is crucial for their expressive power and inductive biases, motivating various theoretical approaches. Some approaches describe network behavior after training through a change in kernel scale from initialization, resulting in a generalization power comparable to a Gaussian process. Conversely, in other approaches training results in the adaptation of the kernel to the data, involving directional changes to the kernel. The relationship and respective strength…

Ocean-E2E: Hybrid Physics-Based and Data-Driven Global Forecasting of Extreme Marine Heatwaves with End-to-End Neural Assimilation
Ruiqi Shu, Yuan Gao, Hao Wu, Ruijian Gou, Yanfei Xiang, Fan Xu, Qingsong Wen, Xian Wu, Xiaomeng Huang
https://arxiv.org/abs/2505.22071

Ocean-E2E: Hybrid Physics-Based and Data-Driven Global Forecasting of Extreme Marine Heatwaves with End-to-End Neural Assimilation
This work focuses on the end-to-end forecast of global extreme marine heatwaves (MHWs), which are unusually warm sea surface temperature events with profound impacts on marine ecosystems. Accurate prediction of extreme MHWs has significant scientific and financial worth. However, existing methods still have certain limitations, especially in the most extreme MHWs. In this study, to address these issues, based on the physical nature of MHWs, we created a novel hybrid data-driven and numerical MH…

Leak of data belonging to 7.4 million Paraguays traced back to infostealers https://therecord.media/data-leak-paraguayan-millions-infostealer

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
- Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future co...
Farouk Mokhtar, Joosep Pata, Dolores Garcia, Eric Wulff, Mengk…

Series B, Episode 04 - Horizon
CALLY: [Enters] Orac, where did you get this information? [Holds up a data card. Avon takes it]
ORAC: I was instructed to obtain anything relating to the planet. The data was obtained by cross-referencing prisoner and execution lists. It is standard procedure.
https://blake.torpidity.n…

Blakes7Bot | Search
Companion site to the Blakes7Bot Twitter, Mastodon, and Bluesky accounts. Automated lines of Blakes 7 dialog. Includes a script search facility.

Generalised Time-Series Analysis of Fault Mechanics Using Explainable AI
Thomas King, Sergio Vinciguerra
https://arxiv.org/abs/2505.21312 https://

Generalised Time-Series Analysis of Fault Mechanics Using Explainable AI
Understanding how faults nucleate and grow is a critical problem in earthquake science and hazard assessment. This study examines fault development in Alzo granite under triaxial pressures ranging from 5 to 40 MPa by applying a Time Delay Neural Network (TDNN) to multi-parameter acoustic emission (AE) data. The TDNN integrates waveform-derived attributes, including peak delay and scattering attenuation, with occurrence-based metrics such as time distributions, Gutenberg-Richter b-values, and sp…

Heute vor 69 Jahren: Am 27. Mai 1956 kam es zum #Atomtest "Yuma". Die Operation #Redwing war eine US-Serie von 17 Atomtestdetonationen von Mai bis Juli 1956. Sie wurden von der Joint Task Force 7 (JTF7) auf den Atollen

From simulations to observations. Methodology and data release of mock TNG50 galaxies at 0.3 < z < 0.7 for WEAVE-StePS
A. Ikhsanova, L. Costantin, A. Pizzella, E. M. Corsini, L. Morelli, F. R. Ditrani, A. Ferr\'e-Mateu, L. Gabarra, M. Gullieuszik, C. P. Haines, A. Iovino, M. Longhetti, A. Mercurio, R. Ragusa, P. S\'anchez-Bl\'azquez, C. Tortora, B. Vulcani, S. Zhou, E. Gafton, F. Pistis

From simulations to observations. Methodology and data release of mock TNG50 galaxies at 0.3 < z < 0.7 for WEAVE-StePS
The new generation of optical spectrographs (i.e., WEAVE, 4MOST, DESI, and WST) offer unprecedented opportunities for statistically studying the star formation histories of galaxies. However, these observations are not easily comparable to predictions from cosmological simulations. Our goal is to build a reference framework for comparing spectroscopic observations with simulations and test tools for deriving stellar population properties of galaxies. We focus on the observational strategy of th…

Using Machine Learning in Analyzing Air Quality Discrepancies of Environmental Impact
Shuangbao Paul Wang, Lucas Yang, Rahouane Chouchane, Jin Guo, Michael Bailey
https://arxiv.org/abs/2506.17319

Using Machine Learning in Analyzing Air Quality Discrepancies of Environmental Impact
In this study, we apply machine learning and software engineering in analyzing air pollution levels in City of Baltimore. The data model was fed with three primary data sources: 1) a biased method of estimating insurance risk used by homeowners loan corporation, 2) demographics of Baltimore residents, and 3) census data estimate of NO2 and PM2.5 concentrations. The dataset covers 650,643 Baltimore residents in 44.7 million residents in 202 major cities in US. The results show that air pollution…

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
- Tunable correlation retention: A statistical method for generating synthetic data
Nicklas J\"averg{\aa}rd, Rainey Lyons, Adrian Muntean, Jonas Forsman
…

Series B, Episode 04 - Horizon
CALLY: [Enters] Orac, where did you get this information? [Holds up a data card. Avon takes it]
ORAC: I was instructed to obtain anything relating to the planet. The data was obtained by cross-referencing prisoner and execution lists. It is standard procedure.
https://blake.torpidity.n…

Blakes7Bot | Search
Companion site to the Blakes7Bot Twitter, Mastodon, and Bluesky accounts. Automated lines of Blakes 7 dialog. Includes a script search facility.

I'm pretty happy with the top side Samtec ARF6 launch simulations now.
S11:
* Red: No cutout, -12 dB at 10 GHz, -7 at 20 GHz, -4.8 at 30 GHz
* Blue: Rectangular ground cutout the size of the pad, -22.5 at 10 GHz, -16.4 at 20 GHz, -11.7 at 30 GHz
* Green: 30um larger than the pad on all sides (1.06 x 0.41 mm), -30.5 at 10 GHz, -22 at 20, -15 at 30.
With the oversized cutout it's more than good enough for QSGMII and 25Gbase-R.
TDR:
* Red: No cutou…

SRG/eROSITA No. 5: Discovery of quasi-periodic eruptions every ~3.7 days from a galaxy at z>0.1
R. Arcodia, P. Baldini, A. Merloni, A. Rau, K. Nandra, J. Chakraborty, A. J. Goodwin, M. J. Page, J. Buchner, M. Masterson, I. Monageng, Z. Arzoumanian, D. Buckley, E. Kara, G. Ponti, M. E. Ramos-Ceja, M. Salvato, K. Gendreau, I. Grotova, M. Krumpe
http…

SRG/eROSITA No. 5: Discovery of quasi-periodic eruptions every ~3.7 days from a galaxy at z>0.1
Quasi-periodic eruptions (QPEs) are repeating soft X-ray bursts from the nuclei of galaxies, tantalizingly proposed to be extreme mass ratio inspirals. Here, we report the discovery of a new galaxy showing X-ray QPEs, the fifth found through a dedicated blind search in the \emph{SRG}/eROSITA all-sky survey data, hereafter named eRO-QPE5. Its QPE duration ($t_{\rm dur}\sim0.6$\,d), recurrence time ($t_{\rm recur}\sim3.7\,$d), integrated energy per eruption ($\sim3.4 \times 10^{47}\,$erg), and bl…

#Transit growth outpaces driving growth in Metro #Vancouver
More people in Metro Vancouver are choosing transit over driving as their primary option to get around, according to new data from TransLink. In 2024, there were 7.7 million more trips on transit, an increase of more than three per cent over 2023, o…

Agent-RewardBench: Towards a Unified Benchmark for Reward Modeling across Perception, Planning, and Safety in Real-World Multimodal Agents
Tianyi Men, Zhuoran Jin, Pengfei Cao, Yubo Chen, Kang Liu, Jun Zhao
https://arxiv.org/abs/2506.21252 https://arxiv.org/pdf/2506.21252 https://arxiv.org/html/2506.21252
arXiv:2506.21252v1 Announce Type: new
Abstract: As Multimodal Large Language Models (MLLMs) advance, multimodal agents show promise in real-world tasks like web navigation and embodied intelligence. However, due to limitations in a lack of external feedback, these agents struggle with self-correction and generalization. A promising approach is to use reward models as external feedback, but there is no clear on how to select reward models for agents. Thus, there is an urgent need to build a reward bench targeted at agents. To address these challenges, we propose Agent-RewardBench, a benchmark designed to evaluate reward modeling ability in MLLMs. The benchmark is characterized by three key features: (1) Multiple dimensions and real-world agent scenarios evaluation. It covers perception, planning, and safety with 7 scenarios; (2) Step-level reward evaluation. It allows for the assessment of agent capabilities at the individual steps of a task, providing a more granular view of performance during the planning process; and (3) Appropriately difficulty and high-quality. We carefully sample from 10 diverse models, difficulty control to maintain task challenges, and manual verification to ensure the integrity of the data. Experiments demonstrate that even state-of-the-art multimodal models show limited performance, highlighting the need for specialized training in agent reward modeling. Code is available at github.
toXiv_bot_toot

Replaced article(s) found for cs.LG. https://arxiv.org/list/cs.LG/new
[7/7]:
- From Data-Driven to Purpose-Driven Artificial Intelligence: Systems Thinking for Data-Analytic Au...
Anadria, Dobbe, Giachanou, Kuiper, Bartels, van Amsterdam, de Troya, Z\"urcher, Oberski
…

[2025-06-27 Fri (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

On World Rainforest Day, the world confronts an unprecedented wave of tropical forest loss https://news.mongabay.com/2025/06/on-world-rainforest-day-the-world-confronts-an-unprecedented-wave-of-tropical-forest-loss/

On World Rainforest Day, the world confronts an unprecedented wave of tropical forest loss
If tropical rainforests are the lungs of the Earth, then 2024 was a year of respiratory failure. Fires, fueled by climate extremes and human recklessness, tore through vast stretches of forest at an unprecedented pace. According to satellite data from the University of Maryland and the World Resources Institute, 6.7 million hectares of tropical primary […]

Heute vor 69 Jahren: Am 27. Mai 1956 kam es zum #Atomtest "Zuni". Die Operation #Redwing war eine US-Serie von 17 Atomtestdetonationen von Mai bis Juli 1956. Sie wurden von der Joint Task Force 7 (JTF7) auf den Atollen

A Global-scale Database of Seismic Phases from Cloud-based Picking at Petabyte Scale
Yiyu Ni, Marine A. Denolle, Amanda M. Thomas, Alex Hamilton, Jannes M\"unchmeyer, Yinzhi Wang, Lo\"ic Bachelot, Chad Trabant, David Mencin
https://arxiv.org/abs/2505.18874

A Global-scale Database of Seismic Phases from Cloud-based Picking at Petabyte Scale
We present the first global-scale database of 4.3 billion P- and S-wave picks extracted from 1.3 PB continuous seismic data via a cloud-native workflow. Using cloud computing services on Amazon Web Services, we launched ~145,000 containerized jobs on continuous records from 47,354 stations spanning 2002-2025, completing in under three days. Phase arrivals were identified with a deep learning model, PhaseNet, through an open-source Python ecosystem for deep learning, SeisBench. To visualize and …

Infant Cry Emotion Recognition Using Improved ECAPA-TDNN with Multiscale Feature Fusion and Attention Enhancement
Junyu Zhou, Yanxiong Li, Haolin Yu
https://arxiv.org/abs/2506.18402

Infant Cry Emotion Recognition Using Improved ECAPA-TDNN with Multiscale Feature Fusion and Attention Enhancement
Infant cry emotion recognition is crucial for parenting and medical applications. It faces many challenges, such as subtle emotional variations, noise interference, and limited data. The existing methods lack the ability to effectively integrate multi-scale features and temporal-frequency relationships. In this study, we propose a method for infant cry emotion recognition using an improved Emphasized Channel Attention, Propagation and Aggregation in Time Delay Neural Network (ECAPA-TDNN) with b…

Series B, Episode 04 - Horizon
ZEN: Data was obtained via Federation cipher. Data also suggests that it is visited annually by a Federation supply freighter.
BLAKE: Carrying what?
ZEN: There is no information.
https://blake.torpidity.net/m/204/40 B7B3

Blakes7Bot | Search
Companion site to the Blakes7Bot Twitter, Mastodon, and Bluesky accounts. Automated lines of Blakes 7 dialog. Includes a script search facility.

A new paper projecting Joshua tree habitat under future climate based on incredibly high-resolution distribution data, from Joshua Tree Genome Project collaborators at USGS. They estimate up to 80% loss of suitable habitat by 2100 under the worst-case climate scenario.
#JoshuaTree #science

Fully Few-shot Class-incremental Audio Classification Using Multi-level Embedding Extractor and Ridge Regression Classifier
Yongjie Si, Yanxiong Li, Jiaxin Tan, Qianhua He, Il-Youp Kwak
https://arxiv.org/abs/2506.18406

Fully Few-shot Class-incremental Audio Classification Using Multi-level Embedding Extractor and Ridge Regression Classifier
In the task of Few-shot Class-incremental Audio Classification (FCAC), training samples of each base class are required to be abundant to train model. However, it is not easy to collect abundant training samples for many base classes due to data scarcity and high collection cost. We discuss a more realistic issue, Fully FCAC (FFCAC), in which training samples of both base and incremental classes are only a few. Furthermore, we propose a FFCAC method using a model which is decoupled into a multi…

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
- Novel computational approaches for ratio distributions with an application to Hake's ratio in eff...
Jozef Han\v{c}, Martina Han\v{c}ov\'a, Dominik Borovsk…

Deaths by domain of protest over 12 Bolivian presidencies.
Same data, same graphing technique, but sorting the categories (by which occurs across the most bins) brings visual order to this data visualization. #dataviz #rstats

The Anatomy of India's Industrial Interdependencies: 7-Digit Product-Level Supply-Use and Input-Output Tables from ASI Data (2016-2022) with a Case Study of the Mobile Phone Sector
Sourish Dutta
https://arxiv.org/abs/2506.13936

The Anatomy of India's Industrial Interdependencies: 7-Digit Product-Level Supply-Use and Input-Output Tables from ASI Data (2016-2022) with a Case Study of the Mobile Phone Sector
Outlines the construction of high-resolution, 7-digit product-level Supply-Use Tables (SUTs) and symmetric Input-Output Tables (IOTs) for the Indian economy, leveraging microdata from the Annual Survey of Industries (ASI) for the period 2016-2022. We delineate a robust methodology encompassing the generation of detailed input and output flows, with a particular focus on the reconciliation of data from registered and unregistered manufacturing sectors through a meticulously developed NPCMS-NIC c…

twitter_higgs: Twitter, Higgs boson (2012)
Data on tweets related to the announcement of the discovery of a new fundamental particle with the features of the Higgs boson on 4th July 2012. Data covers 1-7 July 2012, and includes four types of networks: followers, retweets, replies, and mentions.
This network has 304691 nodes and 563069 edges.
Tags: Social, Online, Weighted, Multilayer

Tesla is sitting on an inventory worth $800M of 10,000 unsold Cybertrucks.
Dealerships have resorted to dumping their glut of the unorthodox pickup trucks in deserted parking lots.
✅ Cybertruck Sales Are So Bad That We Gasped
https://futurism.com/cybertruck-sales-bad

Cybertruck Sales Are So Bad That We Gasped
In the entire first quarter of 2025, Tesla has managed to sell just 7,100 Cybertrucks in the US, according to data from S&P Global Mobility.

Updated line list for the principal isotopologue of carbon monoxide
Vladimir G. Ushakov, Emile S. Medvedev
https://arxiv.org/abs/2506.12216 https://…

Updated line list for the principal isotopologue of carbon monoxide
The line list for the principal isotopologue of CO calculated earlier by the present authors [1, 2] with the irregular dipole-moment function (DMF) is updated with use of the recent high-precision measurements in the 3-0 [3, 4] (Bielska et al. 2022, Hodges et al. 2025) and 7-0 [5] (Balashov et al. 2023) bands. The new data came in contradiction with the experimental data on the 1-0 band [6, 7]. Therefore, we performed fitting several model DMFs to the modified original data set of Meshkov et al…

[2025-06-26 Thu (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Albion EU was offline, now probably not.
During that time I went for a 30 minute pork pie mining round in Believer Tor(T5) on the Albion Online Asia server.
This is some data I noted down on etherpad based on my mining round.
I got 42k silver as you can see. Not much...maybe that's like a joke to you...
#AlbionOnline

Lessons Learned from FPF “Deploying AI Systems” Workshop
https://fpf.org/blog/lessons-learned-from-fpf-deploying-ai-systems-workshop/
@…

Active InSAR monitoring of building damage in Gaza during the Israel-Hamas War
Corey Scher, Jamon Van Den Hoek
https://arxiv.org/abs/2506.14730 https://

Active InSAR monitoring of building damage in Gaza during the Israel-Hamas War
Aerial bombardment of the Gaza Strip beginning October 7, 2023 is one of the most intense bombing campaigns of the twenty-first century, driving widespread urban damage. Characterizing damage over a geographically dynamic and protracted armed conflict requires active monitoring. Synthetic aperture radar (SAR) has precedence for mapping disaster-induced damage with bi-temporal methods but applications to active monitoring during sustained crises are limited. Using interferometric SAR data from S…

Three-dimensional velocity gradient statistics in a mesoscale convection laboratory experiment
Prafulla P. Shevkar, Roshan J. Samuel, Christian Cierpka, J\"org Schumacher
https://arxiv.org/abs/2506.13155

Three-dimensional velocity gradient statistics in a mesoscale convection laboratory experiment
We present three-dimensional velocity gradient statistics from Rayleigh--Bénard convection experiments in a horizontally extended cell of aspect ratio 25, a paradigm for mesoscale convection. The Rayleigh number $Ra$ ranges from $3.7 \times 10^5$ to $4.8 \times 10^6$, and the Prandtl number $Pr$ from 5 to 7.1. Spatio-temporally resolved volumetric data are reconstructed from moderately dense Lagrangian particle tracking measurements. All nine components of the velocity gradient tensor from the…

Blind Source Separation in Biomedical Signals Using Variational Methods
Yasaman Torabi, Shahram Shirani, James P. Reilly
https://arxiv.org/abs/2506.18281 h…

Blind Source Separation in Biomedical Signals Using Variational Methods
This study introduces a novel unsupervised approach for separating overlapping heart and lung sounds using variational autoencoders (VAEs). In clinical settings, these sounds often interfere with each other, making manual separation difficult and error-prone. The proposed model learns to encode mixed signals into a structured latent space and reconstructs the individual components using a probabilistic decoder, all without requiring labeled data or prior knowledge of source characteristics. We …

Robust Moment Identification for Nonlinear PDEs via a Neural ODE Approach
Shaoxuan Chen, Su Yang, Panayotis G. Kevrekidis, Wei Zhu
https://arxiv.org/abs/2506.05245

Robust Moment Identification for Nonlinear PDEs via a Neural ODE Approach
We propose a data-driven framework for learning reduced-order moment dynamics from PDE-governed systems using Neural ODEs. In contrast to derivative-based methods like SINDy, which necessitate densely sampled data and are sensitive to noise, our approach based on Neural ODEs directly models moment trajectories, enabling robust learning from sparse and potentially irregular time series. Using as an application platform the nonlinear Schrödinger equation, the framework accurately recovers govern…

Search for neutron decay into an antineutrino and a neutral kaon in 0.401 megaton-years exposure of Super-Kamiokande
Kamiokande Collaboration, K. Yamauchi, K. Abe, S. Abe, Y. Asaoka, M. Harada, Y. Hayato, K. Hiraide, K. Hosokawa, K. Ieki, M. Ikeda, J. Kameda, Y. Kanemura, Y. Kataoka, S. Miki, S. Mine, M. Miura, S. Moriyama, M. Nakahata, S. Nakayama, Y. Noguchi, G. Pronost, K. Sato, H. Sekiya, K. Shimizu, R. Shinoda, M. Shiozawa, Y. Suzuki, A. Takeda, Y. Takemoto, H. Tanaka, T. Yano, Y.…

Search for neutron decay into an antineutrino and a neutral kaon in 0.401 megaton-years exposure of Super-Kamiokande
We searched for bound neutron decay via $n\to\barν+K^0$ predicted by the Grand Unified Theories in 0.401 Mton$\cdot$years exposure of all pure water phases in the Super-Kamiokande detector. About 4.4 times more data than in the previous search have been analyzed by a new method including a spectrum fit to kaon invariant mass distributions. No significant data excess has been observed in the signal regions. As a result of this analysis, we set a lower limit of $7.8\times10^{32}$ years on the ne…

MUSE-DARK-I: Dark matter halo properties of intermediate-z star-forming galaxies
B. I. Ciocan, N. F. Bouch\'e, J. Fensch, W. Mercier, D. Krajnovi\'c, J. Richard, T. Contini, A. Jeanneau
https://arxiv.org/abs/2506.19721

MUSE-DARK-I: Dark matter halo properties of intermediate-z star-forming galaxies
[Abridged] We analyse the dark matter (DM) halo properties of 127 0.3

The phenomenological renormalization group in neuronal models near criticality
Kaio F. R. Nascimento, Daniel M. Castro, Gustavo G. Cambrainha, Mauro Copelli
https://arxiv.org/abs/2506.14053

The phenomenological renormalization group in neuronal models near criticality
The phenomenological renormalization group (PRG) has been applied to the study of scaleinvariant phenomena in neuronal data, providing evidence for critical phenomena in the brain. However, it remains unclear how reliably these observed signatures indicate genuine critical behavior, as it is not well established how close to criticality a system must be for them to emerge. Here, we rely on neuronal models with known critical points to investigate under which conditions the PRG procedure yields …

Dzisiaj rano popłynęliśmy na wyjątkowy spacerek, który miał szansę w końcu zmęczyć 7,5 letniego huskiego.

#norwegia

1/n

Heute vor 69 Jahren: Am 25. Juni 1956 kam es zum #Atomtest "Dakota". Die Operation #Redwing war eine US-Serie von 17 Atomtestdetonationen von Mai bis Juli 1956. Sie wurden von der Joint Task Force 7 (JTF7) auf den Atollen

[2025-06-25 Wed (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Wildfires push tropical forest loss in Latin America to record highs https://news.mongabay.com/2025/06/wildfires-push-tropical-forest-loss-in-latin-america-to-record-highs/

Wildfires push tropical forest loss in Latin America to record highs
MEXICO CITY — In 2024, six Latin American countries were in the top 10 nations with the highest loss of tropical primary forest, according to recent data from the University of Maryland, U.S. Topping the list were Brazil and Bolivia, in a year that saw the record-breaking loss of 6.7 million hectares (16.6 million acres) […]

Spatio-spectral diarization of meetings by combining TDOA-based segmentation and speaker embedding-based clustering
Tobias Cord-Landwehr, Tobias Gburrek, Marc Deegen, Reinhold Haeb-Umbach
https://arxiv.org/abs/2506.16228

Spatio-spectral diarization of meetings by combining TDOA-based segmentation and speaker embedding-based clustering
We propose a spatio-spectral, combined model-based and data-driven diarization pipeline consisting of TDOA-based segmentation followed by embedding-based clustering. The proposed system requires neither access to multi-channel training data nor prior knowledge about the number or placement of microphones. It works for both a compact microphone array and distributed microphones, with minor adjustments. Due to its superior handling of overlapping speech during segmentation, the proposed pipeline …

twitter_higgs: Twitter, Higgs boson (2012)
Data on tweets related to the announcement of the discovery of a new fundamental particle with the features of the Higgs boson on 4th July 2012. Data covers 1-7 July 2012, and includes four types of networks: followers, retweets, replies, and mentions.
This network has 116408 nodes and 150818 edges.
Tags: Social, Online, Weighted, Multilayer

Transfer entropy for finite data
Alec Kirkley
https://arxiv.org/abs/2506.16215 https://arxiv.org/pdf/2506.16215

Transfer entropy for finite data
Transfer entropy is a widely used measure for quantifying directed information flows in complex systems. While the challenges of estimating transfer entropy for continuous data are well known, it has two major shortcomings that persist even for data of finite cardinality: it exhibits a substantial positive bias for sparse bin counts, and it has no clear means to assess statistical significance. By more precisely accounting for information content in finite data streams, we derive a transfer ent…

Before returning to Albion EU I wanted to get at least full T4 gear on Albion Asia plus Avalonian Pickaxe after using learning points to get to T4 mining. Upgrade costs are part of the data.
As of this moment I give a brief glance at all members in the guild "Mastodon Fediverse 100 Tax":
1/7 online. The 1 that's me of course.
Here are my notes on etherpad(attached image)

Backsolution: A Framework for Solving Inverse Problems via Automatic Differentiation
Koji Kobayashi, Tomi Ohtsuki
https://arxiv.org/abs/2506.13210 https://…

Backsolution: A Framework for Solving Inverse Problems via Automatic Differentiation
We present a simple yet powerful framework for solving inverse problems by leveraging automatic differentiation. Our method is broadly applicable whenever a smooth cost function can be defined near the true solution, and a numerical simulator is available. As a concrete example, we demonstrate that our method can accurately reconstruct the spatial profiles in a conductor from magnetotransport measurements. Even if the given data are insufficient to uniquely determine the profiles, the same fram…

Active Protothrusts and Fluid Highways: Seismic Noise Reveals Hidden Subduction Dynamics in Cascadia
Maleen Kidiwela, Marine A. Denolle, William S. D. Wilcock, Kuan-Fu Feng
https://arxiv.org/abs/2505.15731

Active Protothrusts and Fluid Highways: Seismic Noise Reveals Hidden Subduction Dynamics in Cascadia
Complex interactions between strain accumulation, fault slip, and fluid migration influence shallow subduction zone dynamics. Using a decade of continuous ambient seismic data from Cascadia seafloor observatories, we identified distinct regional variations in subduction dynamics. Northern Cascadia exhibits a fully locked megathrust with persistent strain accumulation, while central Cascadia displays a slow slip event on protothrusts and rapid fluid migration along fault systems in the overridin…

H-QuEST: Accelerating Query-by-Example Spoken Term Detection with Hierarchical Indexing
Akanksha Singh, Yi-Ping Phoebe Chen, Vipul Arora
https://arxiv.org/abs/2506.16751

H-QuEST: Accelerating Query-by-Example Spoken Term Detection with Hierarchical Indexing
Query-by-example spoken term detection (QbE-STD) searches for matching words or phrases in an audio dataset using a sample spoken query. When annotated data is limited or unavailable, QbE-STD is often done using template matching methods like dynamic time warping (DTW), which are computationally expensive and do not scale well. To address this, we propose H-QuEST (Hierarchical Query-by-Example Spoken Term Detection), a novel framework that accelerates spoken term retrieval by utilizing Term Fre…

[2025-06-24 Tue (UTC), 1 new article found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

COSMOS-Web: MIRI Data Reduction and Number Counts at 7.7$\mu$m using JWST
Santosh Harish, Jeyhan S. Kartaltepe, Daizhong Liu, Anton M. Koekemoer, Caitlin M. Casey, Maximilien Franco, Hollis B. Akins, Olivier Ilbert, Marko Shuntov, Nicole E. Drakos, Mike Engesser, Andreas L. Faisst, Ghassem Gozaliasl, Crystal L. Martin, Michaela Hirschmann, Vasily Kokorev, Erini Lambrides, Henry Joy McCracken, Jed McKinney, Louise Paquereau, Jason Rhodes, Brant E. Robertson

COSMOS-Web: MIRI Data Reduction and Number Counts at 7.7$μ$m using JWST
The COSMOS-Web survey is the largest JWST Cycle 1 General Observer program covering a contiguous ~0.54 deg$^2$ area with NIRCam imaging in four broad-band filters and a non-contiguous ~0.2 deg$^2$ with parallel MIRI imaging in a single broad-band filter, F770W. Here we present a comprehensive overview of the MIRI imaging observations, the data reduction procedure, the COSMOS-Web MIRI photometric catalog, and the first data release including the entire COSMOS-Web MIRI coverage. Data reduction is…

72k silver richer after my 30 minute T4 mining round and probably my last one on Albion Asia for now.
As of this moment I give another brief glance at all members in the guild "Mastodon Fediverse 100 Tax":
Still 1/7 online with the 1 also still being me.
Here are my notes on etherpad(attached image) where I've made available some data...
Trivia: Premium cost is 39 M silver.

Algorithm to extract direction in 2D discrete distributions and a continuous Frobenius norm
Jeffrey G. Yepez, Jackson D. Seligman, Max A. A. Dornfest, Brian C. Crow, John G. Learned, Viacheslav A. Li
https://arxiv.org/abs/2506.17360

Algorithm to extract direction in 2D discrete distributions and a continuous Frobenius norm
In this study, we present a novel algorithm for determining directionality in 2D distributions of discrete data. We compare a reference dataset with a known direction to a measured dataset with an unknown direction by the Frobenius norm of the difference (FND) to find the unknown direction. To generalize this concept, we develop a continuous Frobenius norm of the difference (CFND) as a continuous analog of the FND and derive its analytical expression. By relating fitted and normalized 2D Gaussi…

[2025-06-23 Mon (UTC), 2 new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Heute vor 69 Jahren: Am 21. Juni 1956 kam es zum #Atomtest "Inca". Die Operation #Redwing war eine US-Serie von 17 Atomtestdetonationen von Mai bis Juli 1956. Sie wurden von der Joint Task Force 7 (JTF7) auf den Atollen

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
- Exploring blazars through sonification. Visual and auditory insights into multifrequency variability
Gustavo Magallanes-Guij\'on, Sergio Mendoza

Replaced article(s) found for cond-mat.dis-nn. https://arxiv.org/list/cond-mat.dis-nn/new/
[1/1]:
How transformers learn structured data: insights from hierarchical filtering

Factorized RVQ-GAN For Disentangled Speech Tokenization
Sameer Khurana, Dominik Klement, Antoine Laurent, Dominik Bobos, Juraj Novosad, Peter Gazdik, Ellen Zhang, Zili Huang, Amir Hussein, Ricard Marxer, Yoshiki Masuyama, Ryo Aihara, Chiori Hori, Francois G. Germain, Gordon Wichern, Jonathan Le Roux
https://arxiv.org/abs/2506.15…

Factorized RVQ-GAN For Disentangled Speech Tokenization
We propose Hierarchical Audio Codec (HAC), a unified neural speech codec that factorizes its bottleneck into three linguistic levels-acoustic, phonetic, and lexical-within a single model. HAC leverages two knowledge distillation objectives: one from a pre-trained speech encoder (HuBERT) for phoneme-level structure, and another from a text-based encoder (LaBSE) for lexical cues. Experiments on English and multilingual data show that HAC's factorized bottleneck yields disentangled token sets: one…

Redshift ~2.7 is not special
Prajwel Joseph
https://arxiv.org/abs/2506.07774 https://arxiv.org/pdf/2506.07774

Redshift ~2.7 is not special
Galikyan et al. (2025) reported a statistically significant change in galaxy spectral properties at redshift $z$ $\simeq$ 2.7 based on a Kolmogorov analysis of JWST spectroscopic data of galaxies. In this comment, we demonstrate that their result is critically driven by a single outlier in the dataset. This outlier arises from the use of a questionable redshift estimate for one spectrum. When the outlier is removed or the redshift is corrected, the claimed transition at $z$ $\simeq$ 2.7 disappe…

#Blakes7 Series B, Episode 06 - Trial
THANIA: We reserve our opening declaration, sir.
SAMOR: Very well. Enter prosecution data. [A clerk presses some buttons.]
https://blake.torpidity.net/m/206/53

Blakes7Bot | Search
Companion site to the Blakes7Bot Twitter, Mastodon, and Bluesky accounts. Automated lines of Blakes 7 dialog. Includes a script search facility.

Regularized Federated Learning for Privacy-Preserving Dysarthric and Elderly Speech Recognition
Tao Zhong, Mengzhe Geng, Shujie Hu, Guinan Li, Xunying Liu
https://arxiv.org/abs/2506.11069

Regularized Federated Learning for Privacy-Preserving Dysarthric and Elderly Speech Recognition
Accurate recognition of dysarthric and elderly speech remains challenging to date. While privacy concerns have driven a shift from centralized approaches to federated learning (FL) to ensure data confidentiality, this further exacerbates the challenges of data scarcity, imbalanced data distribution and speaker heterogeneity. To this end, this paper conducts a systematic investigation of regularized FL techniques for privacy-preserving dysarthric and elderly speech recognition, addressing differ…

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
The Best Soules Basis for the Estimation of a Spectral Barycentre Network

This https://arxiv.org/abs/2407.19353 has been replaced.
initial toot: https://mastoxiv.page/@arX…

A spring-block theory of feature learning in deep neural networks
Feature-learning deep nets progressively collapse data to a regular low-dimensional geometry. How this emerges from the collective action of nonlinearity, noise, learning rate, and other factors, has eluded first-principles theories built from microscopic neuronal dynamics. We exhibit a noise-nonlinearity phase diagram that identifies regimes where shallow or deep layers learn more effectively and propose a macroscopic mechanical theory that reproduces the diagram and links feature learning acr…

[2025-06-20 Fri (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

BUT System for the MLC-SLM Challenge
Alexander Polok, Jiangyu Han, Dominik Klement, Samuele Cornell, Jan \v{C}ernock\'y, Luk\'a\v{s} Burget
https://arxiv.org/abs/2506.13414

BUT System for the MLC-SLM Challenge
We present a two-speaker automatic speech recognition (ASR) system that combines DiCoW -- a diarization-conditioned variant of Whisper -- with DiariZen, a diarization pipeline built on top of Pyannote. We first evaluate both systems in out-of-domain (OOD) multilingual scenarios without any fine-tuning. In this scenario, DiariZen consistently outperforms the baseline Pyannote diarization model, demonstrating strong generalization. Despite being fine-tuned on English-only data for target-speaker …

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new
[1/1]:
An implementation of neural simulation-based inference for parameter estimation in ATLAS

Liquid and solid layers in a thermal deep learning machine
Gang Huang, Lai Shun Chan, Hajime Yoshino, Ge Zhang, Yuliang Jin
https://arxiv.org/abs/2506.06789

Liquid and solid layers in a thermal deep learning machine
Based on deep neural networks (DNNs), deep learning has been successfully applied to many problems, but its mechanism is still not well understood -- especially the reason why over-parametrized DNNs can generalize. A recent statistical mechanics theory on supervised learning by a prototypical multi-layer perceptron (MLP) on some artificial learning scenarios predicts that adjustable parameters of over-parametrized MLPs become strongly constrained by the training data close to the input/output b…

Pix2Geomodel: A Next-Generation Reservoir Geomodeling with Property-to-Property Translation
Abdulrahman Al-Fakih, Ardiansyah Koeshidayatullah, Nabil A. Saraih, Tapan Mukerji, Rayan Kanfar, Abdulmohsen Alali, SanLinn I. Kaka
https://arxiv.org/abs/2506.17747

Pix2Geomodel: A Next-Generation Reservoir Geomodeling with Property-to-Property Translation
Accurate geological modeling is critical for reservoir characterization, yet traditional methods struggle with complex subsurface heterogeneity, and they have problems with conditioning to observed data. This study introduces Pix2Geomodel, a novel conditional generative adversarial network (cGAN) framework based on Pix2Pix, designed to predict reservoir properties (facies, porosity, permeability, and water saturation) from the Rotliegend reservoir of the Groningen gas field. Utilizing a 7.6 mil…

Embedded Acoustic Intelligence for Automotive Systems
Renjith Rajagopal, Peter Winzell, Sladjana Strbac, Konstantin Lindstr\"om, Petter H\"orling, Faisal Kohestani, Niloofar Mehrzad
https://arxiv.org/abs/2506.11071

Embedded Acoustic Intelligence for Automotive Systems
Transforming sound insights into actionable streams of data, this abstract leverages findings from degree thesis research to enhance automotive system intelligence, enabling us to address road type [1].By extracting and interpreting acoustic signatures from microphones installed within the wheelbase of a car, we focus on classifying road type.Utilizing deep neural networks and feature extraction powered by pre-trained models from the Open AI ecosystem (via Hugging Face [2]), our approach enable…

Robust Unsupervised Adaptation of a Speech Recogniser Using Entropy Minimisation and Speaker Codes
Rogier C. van Dalen, Shucong Zhang, Titouan Parcollet, Sourav Bhattacharya
https://arxiv.org/abs/2506.10653

Robust Unsupervised Adaptation of a Speech Recogniser Using Entropy Minimisation and Speaker Codes
Speech recognisers usually perform optimally only in a specific environment and need to be adapted to work well in another. For adaptation to a new speaker, there is often too little data for fine-tuning to be robust, and that data is usually unlabelled. This paper proposes a combination of approaches to make adaptation to a single minute of data robust. First, instead of estimating the adaptation parameters with cross-entropy on a single error-prone hypothesis or "pseudo-label", this paper pro…

[2025-06-17 Tue (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Better Pseudo-labeling with Multi-ASR Fusion and Error Correction by SpeechLLM
Jeena Prakash, Blessingh Kumar, Kadri Hacioglu, Bidisha Sharma, Sindhuja Gopalan, Malolan Chetlur, Shankar Venkatesan, Andreas Stolcke
https://arxiv.org/abs/2506.11089

Better Pseudo-labeling with Multi-ASR Fusion and Error Correction by SpeechLLM
Automatic speech recognition (ASR) models rely on high-quality transcribed data for effective training. Generating pseudo-labels for large unlabeled audio datasets often relies on complex pipelines that combine multiple ASR outputs through multi-stage processing, leading to error propagation, information loss and disjoint optimization. We propose a unified multi-ASR prompt-driven framework using postprocessing by either textual or speech-based large language models (LLMs), replacing voting or o…

[2025-06-18 Wed (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Fifteen Years of Child-Centered Long-Form Recordings: Promises, Resources, and Remaining Challenges to Validity
Loann Peurey, Marvin Lavechin, Tarek Kunze, Manel Khentout, Lucas Gautheron, Emmanuel Dupoux, Alejandrina Cristia
https://arxiv.org/abs/2506.11075

Fifteen Years of Child-Centered Long-Form Recordings: Promises, Resources, and Remaining Challenges to Validity
Audio-recordings collected with a child-worn device are a fundamental tool in child language research. Long-form recordings collected over whole days promise to capture children's input and production with minimal observer bias, and therefore high validity. The sheer volume of resulting data necessitates automated analysis to extract relevant metrics for researchers and clinicians. This paper summarizes collective knowledge on this technique, providing entry points to existing resources. We als…

Challenges in Automated Processing of Speech from Child Wearables: The Case of Voice Type Classifier
Tarek Kunze, Marianne M\'etais, Hadrien Titeux, Lucas Elbert, Joseph Coffey, Emmanuel Dupoux, Alejandrina Cristia, Marvin Lavechin
https://arxiv.org/abs/2506.11074

Challenges in Automated Processing of Speech from Child Wearables: The Case of Voice Type Classifier
Recordings gathered with child-worn devices promised to revolutionize both fundamental and applied speech sciences by allowing the effortless capture of children's naturalistic speech environment and language production. This promise hinges on speech technologies that can transform the sheer mounds of data thus collected into usable information. This paper demonstrates several obstacles blocking progress by summarizing three years' worth of experiments aimed at improving one fundamental task: V…

[2025-06-16 Mon (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Replaced article(s) found for physics.data-an. https://arxiv.org/list/physics.data-an/new/
[1/1]:
Learning the dynamics of Markovian open quantum systems from experimental data

Unmasking real-world audio deepfakes: A data-centric approach
David Combei, Adriana Stan, Dan Oneata, Nicolas M\"uller, Horia Cucu
https://arxiv.org/abs/2506.09606

Unmasking real-world audio deepfakes: A data-centric approach
The growing prevalence of real-world deepfakes presents a critical challenge for existing detection systems, which are often evaluated on datasets collected just for scientific purposes. To address this gap, we introduce a novel dataset of real-world audio deepfakes. Our analysis reveals that these real-world examples pose significant challenges, even for the most performant detection models. Rather than increasing model complexity or exhaustively search for a better alternative, in this work w…

AC/DC: LLM-based Audio Comprehension via Dialogue Continuation
Yusuke Fujita, Tomoya Mizumoto, Atsushi Kojima, Lianbo Liu, Yui Sudo
https://arxiv.org/abs/2506.10312

AC/DC: LLM-based Audio Comprehension via Dialogue Continuation
We propose an instruction-following audio comprehension model that leverages the dialogue continuation ability of large language models (LLMs). Instead of directly generating target captions in training data, the proposed method trains a model to produce responses as if the input caption triggered a dialogue. This dialogue continuation training mitigates the caption variation problem. Learning to continue a dialogue effectively captures the caption's meaning beyond its surface-level words. As a…

Disentangling Dual-Encoder Masked Autoencoder for Respiratory Sound Classification
Peidong Wei Shiyu Miao Lin Li
https://arxiv.org/abs/2506.10698 https://

Disentangling Dual-Encoder Masked Autoencoder for Respiratory Sound Classification
Deep neural networks have been applied to audio spectrograms for respiratory sound classification, but it remains challenging to achieve satisfactory performance due to the scarcity of available data. Moreover, domain mismatch may be introduced into the trained models as a result of the respiratory sound samples being collected from various electronic stethoscopes, patient demographics, and recording environments. To tackle this issue, we proposed a modified MaskedAutoencoder(MAE) model, named …

[2025-06-13 Fri (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

[2025-06-12 Thu (UTC), no new articles found for physics.data-an Data Analysis, Statistics and Probability]
toXiv_bot_toot

Approaching Dialogue State Tracking via Aligning Speech Encoders and LLMs
\v{S}imon Sedl\'a\v{c}ek, Bolaji Yusuf, J\'an \v{S}vec, Pradyoth Hegde, Santosh Kesiraju, Old\v{r}ich Plchot, Jan \v{C}ernock\'y
https://arxiv.org/abs/2506.08633

Approaching Dialogue State Tracking via Aligning Speech Encoders and LLMs
In this work, we approach spoken Dialogue State Tracking (DST) by bridging the representation spaces of speech encoders and LLMs via a small connector module, with a focus on fully open-sourced and open-data components (WavLM-large, OLMo). We focus on ablating different aspects of such systems including full/LoRA adapter fine-tuning, the effect of agent turns in the dialogue history, as well as fuzzy matching-based output post-processing, which greatly improves performance of our systems on nam…

This https://arxiv.org/abs/2410.00527 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

Wanna hear your voice? A sample is all we need!
Research on audio clue-based target speaker extraction (TSE) has focused on modeling mixtures and reference speech, achieving strong results in English due to abundant datasets. However, cross-lingual properties remain underexplored, as low-resource languages face challenges from limited annotated data and linguistic resources. To bridge this gap, we propose WHYV (Wanna Hear Your Voice), a cross-lingual TSE framework enabling zero-shot adaptation without fine-tuning. WHYV employs a frequency-mo…

A Data-Driven Diffusion-based Approach for Audio Deepfake Explanations
Petr Grinberg, Ankur Kumar, Surya Koppisetti, Gaurav Bharaj
https://arxiv.org/abs/2506.03425

A Data-Driven Diffusion-based Approach for Audio Deepfake Explanations
Evaluating explainability techniques, such as SHAP and LRP, in the context of audio deepfake detection is challenging due to lack of clear ground truth annotations. In the cases when we are able to obtain the ground truth, we find that these methods struggle to provide accurate explanations. In this work, we propose a novel data-driven approach to identify artifact regions in deepfake audio. We consider paired real and vocoded audio, and use the difference in time-frequency representation as th…

Neural Spectral Band Generation for Audio Coding
Woongjib Choi, Byeong Hyeon Kim, Hyungseob Lim, Inseon Jang, Hong-Goo Kang
https://arxiv.org/abs/2506.06732

Neural Spectral Band Generation for Audio Coding
Audio bandwidth extension is the task of reconstructing missing high frequency components of bandwidth-limited audio signals, where bandwidth limitation is a common issue for audio signals due to several reasons, including channel capacity and data constraints. While conventional spectral band replication is a well-established parametric approach to audio bandwidth extension, the SBR usually entails coarse feature extraction and reconstruction techniques, which leads to limitations when process…

AS-ASR: A Lightweight Framework for Aphasia-Specific Automatic Speech Recognition
Chen Bao, Chuanbing Huo, Qinyu Chen, Chang Gao
https://arxiv.org/abs/2506.06566

AS-ASR: A Lightweight Framework for Aphasia-Specific Automatic Speech Recognition
This paper proposes AS-ASR, a lightweight aphasia-specific speech recognition framework based on Whisper-tiny, tailored for low-resource deployment on edge devices. Our approach introduces a hybrid training strategy that systematically combines standard and aphasic speech at varying ratios, enabling robust generalization, and a GPT-4-based reference enhancement method that refines noisy aphasic transcripts, improving supervision quality. We conduct extensive experiments across multiple data mix…

Evaluating the Effectiveness of Pre-Trained Audio Embeddings for Classification of Parkinson's Disease Speech Data
Emmy Postma, Cristian Tejedor-Garcia
https://arxiv.org/abs/2506.02078

Evaluating the Effectiveness of Pre-Trained Audio Embeddings for Classification of Parkinson's Disease Speech Data
Speech impairments are prevalent biomarkers for Parkinson's Disease (PD), motivating the development of diagnostic techniques using speech data for clinical applications. Although deep acoustic features have shown promise for PD classification, their effectiveness often varies due to individual speaker differences, a factor that has not been thoroughly explored in the existing literature. This study investigates the effectiveness of three pre-trained audio embeddings (OpenL3, VGGish and Wav2Vec…

Bridging the Modality Gap: Softly Discretizing Audio Representation for LLM-based Automatic Speech Recognition
Mu Yang, Szu-Jui Chen, Jiamin Xie, John Hansen
https://arxiv.org/abs/2506.05706

Bridging the Modality Gap: Softly Discretizing Audio Representation for LLM-based Automatic Speech Recognition
One challenge of integrating speech input with large language models (LLMs) stems from the discrepancy between the continuous nature of audio data and the discrete token-based paradigm of LLMs. To mitigate this gap, we propose a method for integrating vector quantization (VQ) into LLM-based automatic speech recognition (ASR). Using the LLM embedding table as the VQ codebook, the VQ module aligns the continuous representations from the audio encoder with the discrete LLM inputs, enabling the LLM…

Low-Resource Domain Adaptation for Speech LLMs via Text-Only Fine-Tuning
Yangui Fang, Jing Peng, Xu Li, Yu Xi, Chengwei Zhang, Guohui Zhong, Kai Yu
https://arxiv.org/abs/2506.05671

Low-Resource Domain Adaptation for Speech LLMs via Text-Only Fine-Tuning
Recent advances in automatic speech recognition (ASR) have combined speech encoders with large language models (LLMs) through projection, forming Speech LLMs with strong performance. However, adapting them to new domains remains challenging, especially in low-resource settings where paired speech-text data is scarce. We propose a text-only fine-tuning strategy for Speech LLMs using unpaired target-domain text without requiring additional audio. To preserve speech-text alignment, we introduce a …

PseudoVC: Improving One-shot Voice Conversion with Pseudo Paired Data
Songjun Cao, Qinghua Wu, Jie Chen, Jin Li, Long Ma
https://arxiv.org/abs/2506.01039 h…

PseudoVC: Improving One-shot Voice Conversion with Pseudo Paired Data
As parallel training data is scarce for one-shot voice conversion (VC) tasks, waveform reconstruction is typically performed by various VC systems. A typical one-shot VC system comprises a content encoder and a speaker encoder. However, two types of mismatches arise: one for the inputs to the content encoder during training and inference, and another for the inputs to the speaker encoder. To address these mismatches, we propose a novel VC training method called \textit{PseudoVC} in this paper. …

This https://arxiv.org/abs/2505.21527 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

VietASR: Achieving Industry-level Vietnamese ASR with 50-hour labeled data and Large-Scale Speech Pretraining
Automatic speech recognition (ASR) has made remarkable progress but heavily relies on large-scale labeled data, which is scarce for low-resource languages like Vietnamese. While existing systems such as Whisper, USM, and MMS achieve promising performance, their efficacy remains inadequate in terms of training costs, latency, and accessibility. To address these issues, we propose VietASR, a novel ASR training pipeline that leverages vast amounts of unlabeled data and a small set of labeled data.…

Quality Assessment of Noisy and Enhanced Speech with Limited Data: UWB-NTIS System for VoiceMOS 2024 and Beyond
Marie Kune\v{s}ov\'a
https://arxiv.org/abs/2506.00506

Quality Assessment of Noisy and Enhanced Speech with Limited Data: UWB-NTIS System for VoiceMOS 2024 and Beyond
In this preprint, we present the UWB-NTIS-TTS team's submission to Track 3 of the VoiceMOS 2024 Challenge, the goal of which was to automatically assess the speech quality of noisy and de-noised speech in terms of the ITU-T P.835 metrics of "SIG", "BAK", and "OVRL". Our proposed system, based on wav2vec 2.0, placed among the top systems in the challenge, achieving the best prediction of the BAK scores (background noise intrusiveness), the second-best prediction of the OVRL score (overall audio …