Tootfinder

🏦 NCAA and Power Five conferences agree to deal that will let schools pay athletes
https://www.theguardian.com/sport/article/2024/may/24/ncaa-settlement-conferences-paying-players

This https://arxiv.org/abs/2406.04791 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSD_…

Speaker-Smoothed kNN Speaker Adaptation for End-to-End ASR
Despite recent improvements in End-to-End Automatic Speech Recognition (E2E ASR) systems, the performance can degrade due to vocal characteristic mismatches between training and testing data, particularly with limited target speaker adaptation data. We propose a novel speaker adaptation approach Speaker-Smoothed kNN that leverages k-Nearest Neighbors (kNN) retrieval techniques to improve model output by finding correctly pronounced tokens from its pre-built datastore during the decoding phase. …

LoRA-Whisper: Parameter-Efficient and Extensible Multilingual ASR
Zheshu Song, Jianheng Zhuo, Yifan Yang, Ziyang Ma, Shixiong Zhang, Xie Chen
https://arxiv.org/abs/2406.06619

LoRA-Whisper: Parameter-Efficient and Extensible Multilingual ASR
Recent years have witnessed significant progress in multilingual automatic speech recognition (ASR), driven by the emergence of end-to-end (E2E) models and the scaling of multilingual datasets. Despite that, two main challenges persist in multilingual ASR: language interference and the incorporation of new languages without degrading the performance of the existing ones. This paper proposes LoRA-Whisper, which incorporates LoRA matrix into Whisper for multilingual ASR, effectively mitigating la…

Messengers: Breaking Echo Chambers in Collective Opinion Dynamics with Homophily
Mohsen Raoufi, Heiko Hamann, Pawel Romanczuk
https://arxiv.org/abs/2406.06440

Messengers: Breaking Echo Chambers in Collective Opinion Dynamics with Homophily
Collective estimation manifests computational intelligence emerging from inter-individual local interactions, e.g., by aggregating opinions from neighbors to estimate a quantity. Use cases of collective estimation may include directed motion in physical space, such that agents, for example, have to collectively explore a distributed feature, and collectively agree on a numerical value. In doing so, collectives face several challenges in achieving precise estimations. These challenges exhibit co…

Scintillation Light in SBND: Simulation, Reconstruction, and Expected Performance of the Photon Detection System
SBND Collaboration, P. Abratenko, R. Acciarri, C. Adams, L. Aliaga-Soplin, O. Alterkait, R. Alvarez-Garrote, C. Andreopoulos, A. Antonakis, L. Arellano, J. Asaadi, W. Badgett, S. Balasubramanian, V. Basque, A. Beever, B. Behera, E. Belchior, M. Betancourt, A. Bhat, M. Bishai, A. Blake, B. Bogart, J. Bogenschuetz, D. Brailsford, A. Brandt, S. Brickner, A. Bueno, L. Camilleri,…

Scintillation Light in SBND: Simulation, Reconstruction, and Expected Performance of the Photon Detection System
SBND is the near detector of the Short-Baseline Neutrino program at Fermilab. Its location near to the Booster Neutrino Beam source and relatively large mass will allow the study of neutrino interactions on argon with unprecedented statistics. This paper describes the expected performance of the SBND photon detection system, using a simulated sample of beam neutrinos and cosmogenic particles. Its design is a dual readout concept combining a system of 120 photomultiplier tubes, used for triggeri…

This https://arxiv.org/abs/2406.04791 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSD_…

Speaker-Smoothed kNN Speaker Adaptation for End-to-End ASR
Despite recent improvements in End-to-End Automatic Speech Recognition (E2E ASR) systems, the performance can degrade due to vocal characteristic mismatches between training and testing data, particularly with limited target speaker adaptation data. We propose a novel speaker adaptation approach Speaker-Smoothed kNN that leverages k-Nearest Neighbors (kNN) retrieval techniques to improve model output by finding correctly pronounced tokens from its pre-built datastore during the decoding phase. …

The initial estimate (2 years ago) for keeping California's Diablo Canyon nuclear plant operating was $5.2bn. Now PG&E says it's $8.3bn. And a former state energy commissioner says it's now more like $10 - $11.8bn and "you're pouring your money down a rat hole."
Nuclear costs doubling in two years. And this is my shocked face 😑 (See @…

Critics fear PG&E's aging Diablo Canyon power plant costs may be twice the initial estimates
The cost of keeping the aging Diablo Canyon nuclear plant open five more years could be as much as double what PG&E had first estimated back when a deal was struck two years ago to run it longer to help assure grid reliability, experts say.

"Reuters leaked one of the document’s previous versions. In it, leaders agree on a framework addressing nuclear safety, food security, POW exchanges toward comprehensive peace in Ukraine based on international law. Notably, the draft communiqué didn’t urge Russia to withdraw its troops from Ukraine’s sovereign territory."
**Eighty countries sign communiqué of Ukraine Peace Summit in Switzerland - Euromaidan Press**

Eighty countries sign communiqué of Ukraine Peace Summit in Switzerland - Euromaidan Press
Eighty countries alongside pro-Russian Hungary, Serbia joined four organizations in signing Switzerland's Ukraine peace summit communiqué. South Africa, Saudi Arabia, Thailand, Mexico among those abstained.

This https://arxiv.org/abs/2406.02560 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

Less Peaky and More Accurate CTC Forced Alignment by Label Priors
Connectionist temporal classification (CTC) models are known to have peaky output distributions. Such behavior is not a problem for automatic speech recognition (ASR), but it can cause inaccurate forced alignments (FA), especially at finer granularity, e.g., phoneme level. This paper aims at alleviating the peaky behavior for CTC and improve its suitability for forced alignment generation, by leveraging label priors, so that the scores of alignment paths containing fewer blanks are boosted and …

Transcription-Free Fine-Tuning of Speech Separation Models for Noisy and Reverberant Multi-Speaker Automatic Speech Recognition
William Ravenscroft, George Close, Stefan Goetze, Thomas Hain, Mohammad Soleymanpour, Anurag Chowdhury, Mark C. Fuhs
https://arxiv.org/abs/2406.08914

Transcription-Free Fine-Tuning of Speech Separation Models for Noisy and Reverberant Multi-Speaker Automatic Speech Recognition
One solution to automatic speech recognition (ASR) of overlapping speakers is to separate speech and then perform ASR on the separated signals. Commonly, the separator produces artefacts which often degrade ASR performance. Addressing this issue typically requires reference transcriptions to jointly train the separation and ASR networks. This is often not viable for training on real-world in-domain audio where reference transcript information is not always available. This paper proposes a trans…