Nikita Torgashov PRO

herimor

·

https://herimor.github.io

AI & ML interests

Text-to-Speech, Speaker Recognition

Recent Activity

updated a Space 41 minutes ago

herimor/voxtream2

authored a paper about 18 hours ago

FASTDIAR: Frame-level speaker encoder for Streaming Diarization

updated a model about 21 hours ago

herimor/voxtream2

View all activity

Organizations

Papers 6

arxiv:2610.02941

arxiv:2603.13518

arxiv:2509.15969

arxiv:2407.18223

spaces 3

VoXtream2

Zero-shot TTS with Controllable Speaking-rate

VoXtream

Generate speech from text and an audio prompt

FASTDIAR

Streaming speaker diarization with a 960 ms delay

models 3

herimor/voxtream2

Text-to-Speech • 0.5B • Updated about 21 hours ago • 1.5k • 8

herimor/voxtream2.1

Text-to-Speech • 0.5B • Updated about 21 hours ago • 39

herimor/voxtream

Text-to-Speech • 0.4B • Updated 2 days ago • 90 • 23

datasets 3

herimor/voxtream2-train

Updated 2 days ago • 116

herimor/voxtream2-test

Viewer • Updated 2 days ago • 372 • 52

herimor/voxtream-train-9k

Updated 2 days ago • 511 • 1