Tootfinder

Data-independent Beamforming for End-to-end Multichannel Multi-speaker ASR
Can Cui, Paul Magron, Mostafa Sadeghi, Emmanuel Vincent
https://arxiv.org/abs/2509.10234 https://

Data-independent Beamforming for End-to-end Multichannel Multi-speaker ASR
Automatic speech recognition (ASR) in multichannel, multi-speaker scenarios remains challenging due to ambient noise, reverberation and overlapping speakers. In this paper, we propose a beamforming approach that processes specific angular sectors based on their spherical polar coordinates before applying an end-to-end multichannel, multi-speaker ASR system. This method is data-independent and training-free. We demonstrate that using a group of beamformed signals improves ASR performance compare…

Human Body Segment Volume Estimation with Two RGB-D Cameras
Giulia Bassani, Emilio Maoddi, Usman Asghar, Carlo Alberto Avizzano, Alessandro Filippeschi
https://arxiv.org/abs/2509.10429

Human Body Segment Volume Estimation with Two RGB-D Cameras
In the field of human biometry, accurately estimating the volume of the whole body and its individual segments is of fundamental importance. Such measurements support a wide range of applications that include assessing health, optimizing ergonomic design, and customizing biomechanical models. In this work, we presented a Body Segment Volume Estimation (BSV) system to automatically compute whole-body and segment volumes using only two RGB-D cameras, thus limiting the system complexity. However, …

Are UX evaluation methods truly accessible
Andr\'es Eduardo Fuentes-Cort\'azar, Alejandra Rivera-Hern\'andez, Jos\'e Rafael Rojano-C\'aceres
https://arxiv.org/abs/2508.07620

Are UX evaluation methods truly accessible
Providing an equitable and inclusive user experience (UX) for people with disabilities (PWD) is a central goal of accessible design. In the specific case of Deaf users, whose hearing impairments impact language development and communication, it is essential to consider their specific needs during software evaluation processes. This study aimed to analyze a set of UX evaluation methods suggested in the literature as suitable for Deaf individuals, with the goal of validating their level of access…

Statistical Shape Modelling of the Nasosinusal Anatomy
Michele Bertolini (POLIMI), Silvia Tonghini (POLIMI), Marco Rossoni (POLIMI), Marina Carulli (POLIMI), Susanne Weise (TU Dresden), Christian Jan Baldus (POLIMI), Giorgio Colombo (POLIMI), Monica Bordegoni (POLIMI)
https://arxiv.org/abs/2510.09713…

Statistical Shape Modelling of the Nasosinusal Anatomy
The human nose exhibits a huge variation in shape among individuals. All these variants alter the airflow through the nasal cavity and can impact how we smell odors. To acquire a better understanding of physiological and pathological functioning, it is important to study the effects of these modifications. SSM, or Statistical Shape Modelling, is a widely used methodology for considering morphological differences within a population. In the literature, only a few studies analyze nasal anatomy in…

AugerPrime: Status and first results
David Schmidt (for the Pierre Auger Collaboration)
https://arxiv.org/abs/2508.08056 https://arxiv.org/pdf/2508.08056…

AugerPrime: Status and first results
With the knowledge and statistical precision derived from two decades of measurement, the Pierre Auger Observatory has significantly deepened our understanding of ultra-high-energy cosmic rays while unearthing an increasingly complex astrophysical landscape and exposing tensions with hadronic interaction models. The field now demands the mass of individual cosmic-ray primaries as an observable with an exposure that only the 3000-square-kilometer surface array of the Observatory can provide. Acc…

PARCO: Phoneme-Augmented Robust Contextual ASR via Contrastive Entity Disambiguation
Jiajun He, Naoki Sawada, Koichi Miyazaki, Tomoki Toda
https://arxiv.org/abs/2509.04357 https…

PARCO: Phoneme-Augmented Robust Contextual ASR via Contrastive Entity Disambiguation
Automatic speech recognition (ASR) systems struggle with domain-specific named entities, especially homophones. Contextual ASR improves recognition but often fails to capture fine-grained phoneme variations due to limited entity diversity. Moreover, prior methods treat entities as independent tokens, leading to incomplete multi-token biasing. To address these issues, we propose Phoneme-Augmented Robust Contextual ASR via COntrastive entity disambiguation (PARCO), which integrates phoneme-aware …

“At what point are folks going to try literally any other tactic than condescending rants? Does it matter that LLM apps are at the top of virtually every app store nearly every day because individual people are choosing to download them[…]?”
https://simonwillison.net/2025/Sep/7/anil-dash/

I tried some of the scarier prompts from reddit on the new model and it shut them down straight away. It didn't agree with me or encourage me to keep going.
I don't trust OpenAI as far as I can throw them on ethics but this will probably turn out to be one of their more socially-responsible moves.

Inverse Source Problems for the Time-Fractional Evolution Equation
Rahmonov Askar Ahmadovich
https://arxiv.org/abs/2508.06209 https://arxiv.org/pdf/2508.06…

Inverse Source Problems for the Time-Fractional Evolution Equation
In this paper, we investigate the direct and linear inverse problems of identifying time-dependent and time-independent source terms in a time-fractional diffusion-wave equation, using measured data at an interior point of the time interval. We first establish the existence, uniqueness, and regularity of the solution to the direct problem by employing the spectral method. Then, based on the properties of the direct problem, we study two inverse problems: one involving the recovery of a time-ind…

Acquiescence Bias in Large Language Models
Daniel Braun
https://arxiv.org/abs/2509.08480 https://arxiv.org/pdf/2509.08480

Acquiescence Bias in Large Language Models
Acquiescence bias, i.e. the tendency of humans to agree with statements in surveys, independent of their actual beliefs, is well researched and documented. Since Large Language Models (LLMs) have been shown to be very influenceable by relatively small changes in input and are trained on human-generated data, it is reasonable to assume that they could show a similar tendency. We present a study investigating the presence of acquiescence bias in LLMs across different models, tasks, and languages …