Daniel Duran
banner
simphon.bsky.social
Daniel Duran
@simphon.bsky.social
What?
Reposted by Daniel Duran
Postdoc position in Stuttgart, Germany (TV-L 13, 100%) for 18 months, on authority presuppositions in AI systems with Dr. Agnieszka Faleńska and me. For more information and application info, see here: safety.https://www.ims.uni-stuttgart.de/documents/team/falensaa/aphic_postoc.pdf
www.ims.uni-stuttgart.de
November 24, 2025 at 8:22 AM
Reposted by Daniel Duran
Tomorrow, we will show within the science festival #geniale how AI voice modification can help explaining the subtle differences between voice qualities, expressing personality, age, mood, gender, health, and much more! #trr318, #bielefeld #tts #xAI wissenswerkstadt.de/veranstaltun...
Sag was! | Wissenswerkstadt Bielefeld
Mit Hilfe von KI Stimmeigenschaften erklärbarer machen
wissenswerkstadt.de
November 21, 2025 at 1:15 PM
Reposted by Daniel Duran
📢 Call for papers! Join us in Sydney for #Interspeech2026, 27 September – 1 October 2026. Paper submission portal opens 17 Jan and closes 25 Feb. See you there! interspeech2026.org/en-AU/pages/...
November 20, 2025 at 7:00 AM
Am I evil? Am I likeable?

Need a 10 minutes break? Like Fantasy? Loath it? Take part in our study and help us by rating images of fictional characters here:
bixprag.lili.uni-bielefeld.de/publix/0aSWK...
November 19, 2025 at 10:25 AM
Reposted by Daniel Duran
#CfP alert! 🚨 The call for papers and theme sessions for the next International Conference of the German Cognitive Linguistics Conference (Bielefeld, 31.08.-02.09.2026) is finally out 🎉 www.uni-bielefeld.de/fakultaeten/...
www.uni-bielefeld.de
November 15, 2025 at 3:27 PM
Do you even need visual social cues to trigger speech register effects? Imagining to address a friend vs a policeman is enough to change speaking style (f0 in this case).

Check out our new paper in which we present an analysis of speech data collected with the #Plapper app:
doi.org/10.1121/10.0...
November 11, 2025 at 6:30 PM
Netflix-Doku #Babo — was wäre das unwahrscheinlichste, das ein Film über Haftbefehl auslösen konnte: www.deutschlandfunk.de/haftbefehl-d...
Netflix - "Haftbefehl"-Doku löst Hype um Reinhard Mey aus
Die "Haftbefehl"-Dokumentation auf Netflix hat zahlreiche Fans des Deutsch-Rappers Aykut Anhan an den Liedermacher Reinhard Mey herangeführt. In den Sozialen Medien wird Meys Lied "In meinem Garten" a...
www.deutschlandfunk.de
November 3, 2025 at 10:48 AM
Rund um den Babo-Hype liest man nun oft: „Haftbefehl gehört in die Schulen“
Ganz meine Meinung!
Ich habe da auch schon mal was vorbereitet:
― D. Duran, J. Happes & D. Schumacher (2024). Antisemitismus als „popkulturelles Phänomen“. In N. Hodaie et al. (Hrsg) (Deutsch-) Rap und Gewalt. Beltz Juventa.
November 1, 2025 at 5:33 PM
Immer wieder schön, wenn im Bundesland, in dem man wohnt, ein Feiertag ist – aber im Bundesland, in dem man arbeitet, nicht. Frohes Schaffen!
October 31, 2025 at 6:30 AM
More than 3200 papers? Wicked 🤓
✨We are thrilled to announce that over 3200 papers have been accepted to #EMNLP2025

This includes over 1800 main conference papers and over 1400 papers in findings!

Congratulations to all authors!! 🎉🎉🎉
October 28, 2025 at 5:57 PM
»Lieber Friedrich,
du hast echt bezaubernde Töchter
Wir auch,
aber uns’re hausen in Löchern«
— Eko Fresh. Friedrich. 2025
October 27, 2025 at 6:47 PM
Es muss doch eine bessere Berufsbezeichnung geben 🤔
October 27, 2025 at 12:02 PM
Eine Stunde zu früh zum Termin aus dem Haus gehetzt und erst im Auto den Fehler bemerkt: Check! 🥴

#Zeitumstellung
October 26, 2025 at 8:49 AM
Speech registers are not just about the choice of fancy words or complex linguistic constructions. Speakers also adjust their pronunciation in different situations. Check out our analysis of formal vs informal video interactions using a novel experimental paradigm: doi.org/10.1075/rs.2... #SFB1412
October 24, 2025 at 11:41 AM
This could be the start of an exciting initiative to further research on data-efficient and developmentally plausible LMs. It's a great resource of 45 more or less carefully curated datasets for diverse languages (with 16 fillers which require substantial expansion) ― More contributions are needed!
Preprint alert! We release BabyBabelLM, a multilingual benchmark of developmentally plausible training data. I was responsible for German and Polish data as well as various child-directed wikis. Immensely rewarding project with exceptionally cool co-authors. 🥳🚀
𝐃𝐨 𝐲𝐨𝐮 𝐫𝐞𝐚𝐥𝐥𝐲 𝐰𝐚𝐧𝐭 𝐭𝐨 𝐬𝐞𝐞 𝐰𝐡𝐚𝐭 𝐦𝐮𝐥𝐭𝐢𝐥𝐢𝐧𝐠𝐮𝐚𝐥 𝐞𝐟𝐟𝐨𝐫𝐭 𝐥𝐨𝐨𝐤𝐬 𝐥𝐢𝐤𝐞? 🇨🇳🇮🇩🇸🇪

Here’s the proof! 𝐁𝐚𝐛𝐲𝐁𝐚𝐛𝐞𝐥𝐋𝐌 is the first Multilingual Benchmark of Developmentally Plausible Training Data available for 45 languages to the NLP community 🎉

arxiv.org/abs/2510.10159
October 24, 2025 at 9:50 AM
How to get high quality, naturalistic speech recordings from the same speakers in formal vs informal situations for phonetic analysis? Check out our new paper in Register Studies which introduces a novel experimental framework involving simulated video calls! doi.org/10.1075/rs.2... #SFB1412
October 23, 2025 at 7:13 PM
»Empfinden gehet vor dem Denken, Denken vor dem Sprechen, und Sprechen vor dem Schreiben her. So wie man nicht richtig denken kann, wenn man falſch empfunden hat, ſo kann man auch nicht richtig ſprechen, wenn man nicht richtig gedacht hat, …«
― Johann Christoph Adelung. 1788
October 21, 2025 at 6:57 PM
Nope. Geht gar nicht. 🙉

Neulich auch schon irgendwo gesehen. Im Norden natürlich 😆
Als Norddeutsche ist das für mich homophon, funktioniert also super und kann ich gleich noch in die Folien für Phonetik einbauen 😊
October 20, 2025 at 10:20 AM
Random quote 113:

»Es ist Einigkeit und Recht und Freiheit
Für das deutsche Vaterland
Danach lasst uns alle streben
Brüderlich mit Herz und Hand«
— Manuellsen ft. Haftbefehl. Generation Kanack. 2010.
October 19, 2025 at 10:14 PM
Why is it always reviewer 2? 🤬
October 16, 2025 at 4:12 PM
Reposted by Daniel Duran
In the latest Child Lang. Bulletin, we introduced IASCL's Repository of Research Papers on Underrepresented Languages, a resource designed for easy access and sharing of research on the acquisition of children learning underrepresented languages. Read more here: www.childlanguage.org/underreprese...
October 9, 2025 at 7:02 AM
Reposted by Daniel Duran
Using speech technology to improve communication is super important and @yuanyuanzhang.bsky.social is taking the first important steps with Matthijs Valkering. Incidentally, she shows that a personalised ASR system beats humans in recognising dysarthric speech by a large margin 🤩

#proudpromotor
🇬🇧 “Hey Siri, what’s the weather going to be like?” More and more people are talking to their phones. But speech recognition doesn’t work for everyone. Work by @odettes.bsky.social, @yuanyuanzhang.bsky.social, and Matthijs Valkering now seems to be cautiously changing that.
Computer outperforms humans in speech recognition - Delta
Hey Siri, what's the weather going to be like? More and more people are talking to their phones. But speech recognition doesn't work for everyone. Research by the Multimedia Computing Group now seems ...
delta.tudelft.nl
October 15, 2025 at 11:59 AM
Best Poster Award for #Plapper at #P&P last week!
A first peak at our results shows complex patterns of regional and sociodemographic variation in the phonetic realization of German vowels. More coming soon!
www.leibniz-zas.de/de/plapper
home.uni-leipzig.de/pundp2025
October 14, 2025 at 7:18 AM