🏗️ Building on HF

Yuriy Perezhohin PRO

yuriyvnv

13 5 35

https://scholar.google.com/citations?user=I5uzFtwAAAAJ&hl=en

AI & ML interests

Automatic Speech Recognition, Embeddings, Code Generation, Synthetic Data Generation and Filtering

Recent Activity

posted an update about 17 hours ago

📄 Our paper on synthetic speech augmentation for Estonian and Slovenian ASR is officially out in Machine Learning with Applications. The Estonian and Slovenian Parakeet and Whisper models have been on the Hub since March. This is the peer-reviewed paper behind them, with the full adaptation ladder, ablations, and a large-scale error analysis. Quick recap: we fine-tuned Parakeet-TDT-0.6B-v3 and Whisper-large-v3 on identical data, CommonVoice 17 alone and with ~10 hours per language of synthetic speech (LLM-generated text, LLM-as-judge filtering, TTS). Synthetic speech improves in-domain recognition for every model-language pair, cutting CV17 test WER by 1.3 to 3.7 points over real-data-only fine-tuning. Parakeet, at less than half Whisper's size, reaches the lower WER in every fine-tuned condition. Across 95,070 re-aligned sentence pairs, the architecture largely decides which error types persist, while the adaptation regime mainly changes how many. 📦 Resources - Paper: https://www.sciencedirect.com/science/article/pii/S2666827026002082 - Parakeet ET: https://huggingface.co/yuriyvnv/parakeet-tdt-0.6b-estonian - Parakeet SL: https://huggingface.co/yuriyvnv/parakeet-tdt-0.6b-slovenian - Whisper ET: https://huggingface.co/yuriyvnv/whisper-large-v3-estonian - Whisper SL: https://huggingface.co/yuriyvnv/whisper-large-v3-slovenian - Dataset: https://huggingface.co/datasets/yuriyvnv/synthetic_asr_et_sl - Code: https://github.com/yuriyvnv/TTS-Augmented-ASR If you train ASR for other morphologically rich, low-resource languages, I'd like to hear whether the same error patterns show up. @reach-vb @ylacombe @hf-audio #speech #asr #syntheticdata #lowresource #estonian #slovenian

updated a model about 17 hours ago

yuriyvnv/whisper-large-v3-estonian

updated a model about 17 hours ago

yuriyvnv/parakeet-tdt-0.6b-estonian

View all activity

Organizations

upvoted a paper 7 months ago

HM-RAG: Hierarchical Multi-Agent Multimodal Retrieval Augmented Generation

Paper • 2504.12330 • Published Apr 13, 2025 • 1

upvoted a collection about 1 year ago

Whisper Models Dutch Language

Collection

This repo contains Dutch Whisper models finetuned on CV and other synthetic data, with different filtering options • 11 items • Updated Sep 16, 2025 • 3

upvoted 2 collections over 1 year ago

Whisper Models Portuguese Language

Collection

This Repo contains Whisper models trained on subsets of data like Common Voice 17(CV_17), Synthetic(Generated by OpenAI) + CV17 and Synthetic Only. • 13 items • Updated Mar 2 • 2

Parallel Sentences Datasets

Collection

These datasets all have "english" and "non_english" columns for numerous datasets. They can be used to make embedding models multilingual. • 14 items • Updated Aug 18 • 23

upvoted a paper about 2 years ago

Seamless: Multilingual Expressive and Streaming Speech Translation

Paper • 2312.05187 • Published Dec 8, 2023 • 14

Yuriy Perezhohin PRO

AI & ML interests

Recent Activity

Organizations

yuriyvnv's activity