Speech Technology
Kanalga Telegramāda oātish
Ko'proq ko'rsatish
1 726
Obunachilar
+124 soatlar
+117 kun
+3630 kun
Ma'lumot yuklanmoqda...
O'xshash kanallar
Taglar buluti
Kirish va chiqish esdaliklari
---
---
---
---
---
---
Obunachilarni jalb qilish
Sentabr '26
Sentabr '26
+36
1 kanalda
Avgust '26
+69
1 kanalda
Get PRO
Iyul '26
+55
0 kanalda
Get PRO
Iyun '26
+29
1 kanalda
Get PRO
May '26
+27
0 kanalda
Get PRO
Aprel '26
+36
0 kanalda
Get PRO
Mart '26
+20
1 kanalda
Get PRO
Fevral '26
+38
0 kanalda
Get PRO
Yanvar '26
+23
0 kanalda
Get PRO
Dekabr '25
+16
0 kanalda
Get PRO
Noyabr '25
+25
0 kanalda
Get PRO
Oktabr '25
+26
1 kanalda
Get PRO
Sentabr '25
+45
0 kanalda
Get PRO
Avgust '25
+18
0 kanalda
Get PRO
Iyul '25
+18
0 kanalda
Get PRO
Iyun '25
+28
1 kanalda
Get PRO
May '25
+34
0 kanalda
Get PRO
Aprel '25
+34
1 kanalda
Get PRO
Mart '25
+40
1 kanalda
Get PRO
Fevral '25
+36
0 kanalda
Get PRO
Yanvar '25
+20
1 kanalda
Get PRO
Dekabr '24
+56
2 kanalda
Get PRO
Noyabr '24
+37
1 kanalda
Get PRO
Oktabr '24
+137
3 kanalda
Get PRO
Sentabr '24
+47
2 kanalda
Get PRO
Avgust '24
+51
1 kanalda
Get PRO
Iyul '24
+62
2 kanalda
Get PRO
Iyun '24
+34
1 kanalda
Get PRO
May '24
+41
1 kanalda
Get PRO
Aprel '24
+33
1 kanalda
Get PRO
Mart '24
+55
2 kanalda
Get PRO
Fevral '24
+56
2 kanalda
Get PRO
Yanvar '24
+77
1 kanalda
Get PRO
Dekabr '23
+96
3 kanalda
Get PRO
Noyabr '23
+47
2 kanalda
Get PRO
Oktabr '23
+37
1 kanalda
Get PRO
Sentabr '23
+23
0 kanalda
Get PRO
Avgust '23
+24
0 kanalda
Get PRO
Iyul '23
+34
0 kanalda
Get PRO
Iyun '23
+20
0 kanalda
Get PRO
May '23
+79
0 kanalda
Get PRO
Aprel '23
+51
0 kanalda
Get PRO
Mart '23
+15
0 kanalda
Get PRO
Fevral '23
+6
0 kanalda
Get PRO
Yanvar '23
+11
0 kanalda
Get PRO
Dekabr '22
+12
0 kanalda
Get PRO
Noyabr '22
+22
0 kanalda
Get PRO
Oktabr '22
+14
0 kanalda
Get PRO
Sentabr '22
+25
0 kanalda
Get PRO
Avgust '22
+53
0 kanalda
Get PRO
Iyul '22
+9
0 kanalda
Get PRO
Iyun '22
+16
0 kanalda
Get PRO
May '22
+14
0 kanalda
Get PRO
Aprel '22
+26
0 kanalda
Get PRO
Mart '22
+9
0 kanalda
Get PRO
Fevral '22
+8
0 kanalda
Get PRO
Yanvar '22
+10
0 kanalda
Get PRO
Dekabr '21
+16
0 kanalda
Get PRO
Noyabr '21
+11
0 kanalda
Get PRO
Oktabr '21
+16
0 kanalda
Get PRO
Sentabr '21
+3
0 kanalda
Get PRO
Avgust '21
+24
0 kanalda
Get PRO
Iyul '21
+13
0 kanalda
Get PRO
Iyun '21
+16
0 kanalda
Get PRO
May '21
+220
0 kanalda
| Sana | Obunachilarni jalb qilish | Esdaliklar | Kanallar | |
| 14 Sentabr | +2 | |||
| 13 Sentabr | +1 | |||
| 12 Sentabr | +3 | |||
| 11 Sentabr | +6 | |||
| 10 Sentabr | 0 | |||
| 09 Sentabr | 0 | |||
| 08 Sentabr | +6 | |||
| 07 Sentabr | +2 | |||
| 06 Sentabr | +4 | |||
| 05 Sentabr | +1 | |||
| 04 Sentabr | +5 | |||
| 03 Sentabr | +3 | |||
| 02 Sentabr | +2 | |||
| 01 Sentabr | +1 |
Kanal postlari
2.1k parameters VAD
https://github.com/AydinAdnan/PulseVAD
reimplementation of KiloVAD
https://arxiv.org/abs/2607.25870
| 2 | 10000xRT for Zipformer on A100, 18000xRT on H200 with specialized CUDA tricks
https://github.com/SoundsGoodAI/fast-gpu-asr | 350 |
| 3 | Just another reminder RL is important
https://x.com/kadirnardev/status/2098425551203475665 | 334 |
| 4 | Kokoro distill 82M -> 7M params
https://huggingface.co/spaces/oddadmix/Kokoro-7M-Distill-Demo | 384 |
| 5 |  https://x.com/unilightwf/status/2098261200480174123
https://arxiv.org/abs/2603.14328
Cross-lingual cloning TTS is still a big problem, some accent metrics demo interesting results
Also worth checking
https://iwslt.org/2026/voice-cloning
with some useful data
https://huggingface.co/datasets/ymoslem/acl-6060 | 433 |
| 6 | My friend Miro recommended me Orukeet model
https://github.com/Oruk-AI/orukeet
https://arxiv.org/abs/2609.10054
it is indeed a good parakeet improvement, about 10% better. Interesting that people left scaling and return back to in-depth architecture analysis.
Oruk.AI does some other nice things, for example a visualization of emotion representation in different layers of speech models
https://x.com/OrukLabs/status/2073457781018087473
https://oruk.ai/research/how-models-represent-speech | 477 |
| 7 | Interesting TTS, claims 14ms TTFA
https://github.com/X-Square-Robot/X2Streaming-TTS | 507 |
| 8 |  Suplime results on Russian telephony data, good results actually second after Diarizen Large
A bit slow though | 794 |
| 9 | Pyannote improved
https://github.com/rewayai/suplime/ | 715 |
| 10 | https://huggingface.co/tencent/AuK
AuKĀ is a 1.5B foundation model for speech generation and editing. Trained on millions of hours of diverse audio data, AuK supports zero-shot and instruction-based TTS, content and acoustic editing, paralinguistic editing, speech enhancement, and source separation through a unified natural-language instruction interface. | 611 |
| 11 | ParsVoice, the largest open-source Persian speech dataset, along with a TTS model and an open-source processing pipeline is released.
The paper has also been accepted as a main conference paper at EMNLP 2026.
Paper: https://arxiv.org/abs/2510.10774
Dataset: https://huggingface.co/datasets/MohammadJRanjbar/ParsVoice
TTS model: https://huggingface.co/MohammadJRanjbar/ParsVoice-XTTS
Code & pipeline: https://github.com/MohammadJRanjbar/ParsVoice | 783 |
| 12 | Bodhan AI together with AI4Bharat recently released a great update on Indic ASR
https://bodhan.ai/research/blogs/indic-transcribe | 708 |
| 13 | Scicom from Malaysia tries Ascend 910B3
https://github.com/Scicom-AI-Enterprise-Organization/TTS-API-Neucodec/blob/main/ASCEND_910B3_PRECISION_REPORT.md | 690 |
| 14 | https://arxiv.org/abs/2609.01246v1
Ready to Speak: Aligning LLMs for TTS-Friendly Text Generation
Thibaut Thonet,Ā Jos Rozen,Ā Laurent Besacier
Current Large Language Models (LLMs) are primarily optimized for written text, often producing outputs that are grammatically correct and helpful yet poorly suited for spoken delivery via Text-to-Speech (TTS). In this work, we study how to make LLMs natively generate TTS-friendly text, which we frame as a preference alignment problem: instead of relying on downstream rewriting modules, we directly align LLMs to generate text optimized for spoken delivery. We introduce two preference datasets spanning different target domains, CORA and Recipe, which contain paired TTS-friendly and TTS-unfriendly responses. We further propose an evaluation suite combining a pattern-based heuristic metric, a TTSāASR evaluation pipeline, and a MUSHRA listening study with human judges. Our experiments compare the recently proposed Feature-aware Sampling and Tuning (FaST) framework -- leveraging interpretable features instead of a black-box reward model -- against an array of alignment baselines on the TTS-friendly generation task. Notably, we found that FaST achieves the best overall tradeoff between TTS-friendliness and helpfulness across various settings. We also identified a strong correlation between our different metrics, highlighting the ability to reliably assess TTS-friendliness via an efficient heuristic. | 1 170 |
| 15 | Quite an obvious but so ignored by industry before. There is certainly no need to clone from 3 seconds
https://www.linkedin.com/posts/soniox_soniox-texttospeech-voiceai-activity-7501594413660401664-p9Cq | 660 |
| 16 | https://www.nature.com/articles/srep12881
Human starts to plan answer 2 seconds before question end | 952 |
| 17 | We know human scores are useless but anyway
https://x.com/datapointai/status/2094829412625654141
today, we're releasing the largest open-source human audio preferences dataset, focused on the customer support use-case
- 300K+ annotations by real people
- 15 SOTA TTS models ranked (Sonic 3.6, Grok TTS, Simba 3.2, Eleven Labs v3)
- 8 categories (IVR menus, empathy, escalations, refunds etc)
dataset + benchmark + frontier plot below: | 896 |
| 18 | https://turnbench.sesame.com/ | 923 |
| 19 | You can download really huge datasets these days
https://x.com/ahochlehnert/status/2092648676829413778
LAION-BVD: a 10-million-hour open video dataset for multimodal pre-training. - 1.3B video URLs from CommonCrawl - 80M downloaded videos - 10M video hours - 55M captioned clips - 300M frame-caption pairs
This repository containsĀ 1.7 million audio clipsĀ taken from BVD-V-55M and sampled for uniqueness of the source video, so that the subset maximises source diversity rather than clip count. Each clip comes with a caption, its language, and the timestamps locating it in the source video.
https://huggingface.co/datasets/laion/BVD-A-1.7M | 1 305 |
| 20 | https://huggingface.co/BreezeBlue/Breeze-TTS-2
https://breezeblue.ai/breeze-tts-2
English/Chinese only but really good quality | 1 032 |
