Tootfinder

LLM-based speaker diarization correction: A generalizable approach
Georgios Efstathiadis, Vijay Yadav, Anzar Abbas
https://arxiv.org/abs/2406.04927 https:/…

LLM-based speaker diarization correction: A generalizable approach
Speaker diarization is necessary for interpreting conversations transcribed using automated speech recognition (ASR) tools. Despite significant developments in diarization methods, diarization accuracy remains an issue. Here, we investigate the use of large language models (LLMs) for diarization correction as a post-processing step. LLMs were fine-tuned using the Fisher corpus, a large dataset of transcribed conversations. The ability of the models to improve diarization accuracy in a holdout d…

With Awakari getting the spotlight again recently, I think it's time to remind everyone.
Unless you set your posts to "followers only", there's no safety from scrapers, AI, manual collection, etc.
Even more so, your "Public" posts can EASILY be grabbed if you're on Mastodon through your RSS feed.
There's 20k Fediverse instances out there and when you create a Fedi instance, it federates by default unless specifically specified. There is no guaranteed way to verify what each instance is doing with your posts.
Fediverse is not some kind of "bastion of privacy" like some seem to think.
If you want your posts to be private and almost guaranteed unscrapable, set your posts to "Followers Only".
Thanks for coming to my FED talk.
#Fediverse #Fedi #ActivityPub #Mastodon #Privacy #Decentralization #Federation

Charlamagne tha God says Democrats lack authenticity and need to talk 'like real people' (Asher Notheis/Washington Examiner)
https://www.washingtonexaminer.com/news/3043846/charlamagne-tha-god-says-democrats-lack-authenticity-need-talk-like-real-people/
http://www.memeorandum.com/240613/p65#a240613p65

#Adobe "rapes" the digital rights of all of their customers and steals everything what you upload to their tools for their own purpose, taking all of your tools and current data as a hostage until you agree:
https://www.

I found this gem of a slide from @… previous talk about #azure costs and planning. #dotnet

Rigorous Safety Analysis and Design of ADAS and ADS: Implications on Tools
Juan Pimentel
https://arxiv.org/abs/2406.08350 https://arx…

Rigorous Safety Analysis and Design of ADAS and ADS: Implications on Tools
Currently, a major concern is the insufficient level of safety offered by commercial automated vehicles and/or services such self-driving vehicles, self-driving trucks, and robotaxis. Unfortunately, stakeholders do not agree on definitions and characterizations of what is meant by safety of automated vehicles including how to measure it and how to design for it. This paper sheds some light into the answers to important questions about the safety of automated vehicles. In addition, we identify r…

The Pristine survey XXV: The Galactic underdogs -- Dynamic tales of a Milky Way metal-poor population
Isaure Gonz\'alez Rivera de La Vernhe, Vanessa Hill, Georges Kordopatis, Felipe Gran, Emma Fern\'andez-Alvar, Anke Ardern-Arentsen, Guillaume F. Thomas, Federico Sestito, Camila Navarrete, Nicolas F. Martin, Else Starkenburg, Akshara Viswanathan, Giuseppina Battaglia, Kim A. Venn, Sara Vitali

The Pristine survey XXV: The Galactic underdogs -- Dynamic tales of a Milky Way metal-poor population
Through the chemodynamical characterisation of metal-poor stars, one can efficiently probe the early history of the Milky Way. We aim at decontaminating a sample of $\sim$ 3M giant stars with Gaia DR3 XP-based \textit{Pristine-Gaia} metallicities, to investigate a subset of very metal-poor stars ([Fe/H] < -1.7) with disc-like orbits. We construct a statistically robust sample of $\sim$ 36 000 very metal-poor giants, using APOGEE and LAMOST to estimate and remove contamination from high $V_ϕ$ s…