README
🚀
Ojin AI
OjinWorld
AI & ML interests
Real-time voice and face models. Interested in turn-taking and endpointing, streaming ASR, expressive TTS, and talking-head generation. Berlin.
Recent Activity
new activity 10 days ago
openbmb/VoxCPM2:Speed issue in results. new activity 10 days ago
Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice:Question: real-time audio on mobile? new activity 10 days ago
nvidia/parakeet-tdt-0.6b-v3:Questions about streaming with Parakeet and TDT merging methodsOrganizations
Speed issue in results.
4
#3 opened 6 months ago
by 
Hanswalter
Question: real-time audio on mobile?
1
#50 opened 3 months ago
by 
3morixd
Questions about streaming with Parakeet and TDT merging methods
👍 1
3
#13 opened about 1 year ago
by 
alexandreacff
updated a
Space 16 days ago
published a
Space 16 days ago
Thank you to everyone who developed this model.
❤️ 7
1
#7 opened 5 months ago
by 
abbas78
Feedback: Japanese transcription feedback for future improvements
👍❤️ 1
1
#22 opened about 2 months ago
by 
ShahzaibAli4076
Thank You for Open-Sourcing Your Model & Feedback
🧠👍 5
4
#8 opened over 1 year ago
by 
rameshch
Works great in ComfyUI
❤️ 1
3
#26 opened over 1 year ago
by 
Llynara
Model hallucinations/Artifacts Issue
1
#53 opened 5 months ago
by 
sandhu447
Farsi with Hindi Accent issue
1
#15 opened 6 months ago
by 
devops724
Question about reproducing the reported Nemotron 3.5 ASR benchmark results
3
#27 opened 2 months ago
by 
HoussamLCHMT
Benchmark: FunASR 170x realtime vs Whisper 13x — non-autoregressive alternative
3
#233 opened 4 months ago
by 
langgz
Qwen3-ASR ComfyUI integration
👍 3
1
#11 opened 8 months ago
by 
1038lab
Thank You for Qwen3-TTS — ComfyUI Custom Node Integration
👍 3
1
#14 opened 8 months ago
by 1038lab
really nice work!
#1 opened 30 days ago
by OjinWorld
liked a
Space 30 days ago
Update index.html
1
#1 opened about 1 year ago
by 
fernandopruebas