Eleanor Chodroff
@echodroff.bsky.social
Cognitive scientist, linguist, phonetician at the University of Zurich Dept. of Computational Linguistics
Reposted by Eleanor Chodroff
A memorial service, followed by a reception, will take place on Friday, 14 November at 5:00 PM in the Aula (KOL-G-201) of the main building of the University of Zurich (Rämistrasse 71, 8006 Zurich).
October 7, 2025 at 7:54 PM
A memorial service, followed by a reception, will take place on Friday, 14 November at 5:00 PM in the Aula (KOL-G-201) of the main building of the University of Zurich (Rämistrasse 71, 8006 Zurich).
The differences are small but consistent in direction, supporting a biomechanical account critically tied to uniform phonetic targets across vowels. At the same time, variation in effect size across languages suggests speakers differ in how strongly this uniformity is realized
October 2, 2025 at 4:46 PM
The differences are small but consistent in direction, supporting a biomechanical account critically tied to uniform phonetic targets across vowels. At the same time, variation in effect size across languages suggests speakers differ in how strongly this uniformity is realized
Our findings:
📉 Clear crosslinguistic bias—high vowels are shorter than low vowels
➡️ But no systematic difference between high front vs. high back vowels
📉 Clear crosslinguistic bias—high vowels are shorter than low vowels
➡️ But no systematic difference between high front vs. high back vowels
October 2, 2025 at 4:46 PM
Our findings:
📉 Clear crosslinguistic bias—high vowels are shorter than low vowels
➡️ But no systematic difference between high front vs. high back vowels
📉 Clear crosslinguistic bias—high vowels are shorter than low vowels
➡️ But no systematic difference between high front vs. high back vowels
Previous explanations have focused on two explanations:
🗣️Automatic accounts
👂Speaker control
As a novel contribution, we reinterpret intrinsic vowel duration as a statistical universal emerging from the competing pressures of target uniformity and enhancement
🗣️Automatic accounts
👂Speaker control
As a novel contribution, we reinterpret intrinsic vowel duration as a statistical universal emerging from the competing pressures of target uniformity and enhancement
October 2, 2025 at 4:46 PM
Previous explanations have focused on two explanations:
🗣️Automatic accounts
👂Speaker control
As a novel contribution, we reinterpret intrinsic vowel duration as a statistical universal emerging from the competing pressures of target uniformity and enhancement
🗣️Automatic accounts
👂Speaker control
As a novel contribution, we reinterpret intrinsic vowel duration as a statistical universal emerging from the competing pressures of target uniformity and enhancement
We provide a large-scale crosslinguistic corpus analysis of intrinsic vowel duration – the observation that high vowels (like /i/ or /u/) tend to be shorter than low vowels (like /a/)
Our dataset:
✅ 60+ languages
✅ 16 language families
✅ Thousands of speakers
Our dataset:
✅ 60+ languages
✅ 16 language families
✅ Thousands of speakers
October 2, 2025 at 4:46 PM
We provide a large-scale crosslinguistic corpus analysis of intrinsic vowel duration – the observation that high vowels (like /i/ or /u/) tend to be shorter than low vowels (like /a/)
Our dataset:
✅ 60+ languages
✅ 16 language families
✅ Thousands of speakers
Our dataset:
✅ 60+ languages
✅ 16 language families
✅ Thousands of speakers
✅Similarity scores: huggingface.co/datasets/pac...
📄Paper: www.isca-archive.org/interspeech_...
💻Code: github.com/pacscilab/CV...
💫This was joint work with @mzhang89.bsky.social, Aref Farhadipour, Annie Baker, Jiachen Ma, and Bogdan Pricop
📄Paper: www.isca-archive.org/interspeech_...
💻Code: github.com/pacscilab/CV...
💫This was joint work with @mzhang89.bsky.social, Aref Farhadipour, Annie Baker, Jiachen Ma, and Bogdan Pricop
pacscilab/VoxCommunis at main
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
August 29, 2025 at 10:25 AM
✅Similarity scores: huggingface.co/datasets/pac...
📄Paper: www.isca-archive.org/interspeech_...
💻Code: github.com/pacscilab/CV...
💫This was joint work with @mzhang89.bsky.social, Aref Farhadipour, Annie Baker, Jiachen Ma, and Bogdan Pricop
📄Paper: www.isca-archive.org/interspeech_...
💻Code: github.com/pacscilab/CV...
💫This was joint work with @mzhang89.bsky.social, Aref Farhadipour, Annie Baker, Jiachen Ma, and Bogdan Pricop
Pairs below this were more likely perceived as different speakers and above, as the same speaker. Of course there’s no ground truth, so you can also choose your own threshold
The similarity scores, paper, and code can be found at the below links
Happy data cleaning 😊
The similarity scores, paper, and code can be found at the below links
Happy data cleaning 😊
August 29, 2025 at 10:25 AM
Pairs below this were more likely perceived as different speakers and above, as the same speaker. Of course there’s no ground truth, so you can also choose your own threshold
The similarity scores, paper, and code can be found at the below links
Happy data cleaning 😊
The similarity scores, paper, and code can be found at the below links
Happy data cleaning 😊
We ran automatic speaker verification (ResNet-293 trained w/ multilingual VoxBlink2) to obtain similarity scores among files for each client ID. Based on previous thresholds and a perceptual evaluation, we found an optimal threshold of ~0.35–0.40 for same vs diff speakers
August 29, 2025 at 10:25 AM
We ran automatic speaker verification (ResNet-293 trained w/ multilingual VoxBlink2) to obtain similarity scores among files for each client ID. Based on previous thresholds and a perceptual evaluation, we found an optimal threshold of ~0.35–0.40 for same vs diff speakers