ONNX graphs for audiosronnx: denoising, voice isolation, restoration and bandwidth extension to 48 kHz.
AI & ML interests
None defined yet.
Recent Activity
View all activity
Decision models for European Portuguese: Kev fine-tuned on typed decisions built from the AMALIA data.
ONNX Arabic diacritization models used by text2tashkeel: the rawi family plus mirrored bilstm and libtashkeel.
datasets used to train phonemizers
models trained for https://github.com/TigreGotico/phoonnx
datasets ready for training TTS models
Synthetic wake-word audio for English wake words: TTS and voice-cloned clips, one dataset per word.
Synthetic wake-word audio for Portuguese wake words: TTS and voice-cloned clips, one dataset per word.
negative samples for wake word training
ONNX exports powering the vconnx voice-conversion library: one repo per engine, with parity reports and provenance.
-
TigreGotico/model2vec-ATC-role-classification-bert-base-uwb-atcc
7.62M • Updated • 4 -
TigreGotico/model2vec-ATC-role-classification-potion-2M
1.95M • Updated • 7 -
TigreGotico/model2vec-ATC-role-classification-potion-4M
3.84M • Updated • 8 -
TigreGotico/model2vec-ATC-role-classification-potion-8M
7.62M • Updated • 5
Decision models for Arabic, MSA and dialects: Kev fine-tuned on typed decisions from 80 public datasets.
Various models in onnx format meant for audio feature extraction
taxonomy, datasets and baseline models for question type classification
ONNX weights for the speech-quality metrics used by speechonnxmetrics.
Meaning-labelled datasets for European-Portuguese heterophonic homograph disambiguation (bifonia): synthetic train/test and a real-world OOD set.
Voice-converted TTS training datasets
Synthetic audio of the phrase that takes a voice assistant out of sleep mode, one dataset per language.
Synthetic wake-word audio for Spanish wake words: TTS and voice-cloned clips, one dataset per word.
CoreML conversions of community VAD models (fp32 + int8) targeting iOS 17+.
-
TigreGotico/atc-role-classification-synthetic
Viewer • Updated • 10.2k • 22 -
TigreGotico/atc-role-classification-synthetic-2
Viewer • Updated • 298k • 15 -
TigreGotico/atc-role-classification-synthetic-3
Viewer • Updated • 5.08k • 11 -
jacktol/atc-pilot-speaker-role-classification-dataset
Viewer • Updated • 6.89k • 25 • 2
ONNX graphs for audiosronnx: denoising, voice isolation, restoration and bandwidth extension to 48 kHz.
Decision models for Arabic, MSA and dialects: Kev fine-tuned on typed decisions from 80 public datasets.
Decision models for European Portuguese: Kev fine-tuned on typed decisions built from the AMALIA data.
Various models in onnx format meant for audio feature extraction
ONNX Arabic diacritization models used by text2tashkeel: the rawi family plus mirrored bilstm and libtashkeel.
taxonomy, datasets and baseline models for question type classification
datasets used to train phonemizers
ONNX weights for the speech-quality metrics used by speechonnxmetrics.
models trained for https://github.com/TigreGotico/phoonnx
Meaning-labelled datasets for European-Portuguese heterophonic homograph disambiguation (bifonia): synthetic train/test and a real-world OOD set.
datasets ready for training TTS models
Voice-converted TTS training datasets
Synthetic wake-word audio for English wake words: TTS and voice-cloned clips, one dataset per word.
Synthetic audio of the phrase that takes a voice assistant out of sleep mode, one dataset per language.
Synthetic wake-word audio for Portuguese wake words: TTS and voice-cloned clips, one dataset per word.
Synthetic wake-word audio for Spanish wake words: TTS and voice-cloned clips, one dataset per word.
negative samples for wake word training
CoreML conversions of community VAD models (fp32 + int8) targeting iOS 17+.
ONNX exports powering the vconnx voice-conversion library: one repo per engine, with parity reports and provenance.
-
TigreGotico/atc-role-classification-synthetic
Viewer • Updated • 10.2k • 22 -
TigreGotico/atc-role-classification-synthetic-2
Viewer • Updated • 298k • 15 -
TigreGotico/atc-role-classification-synthetic-3
Viewer • Updated • 5.08k • 11 -
jacktol/atc-pilot-speaker-role-classification-dataset
Viewer • Updated • 6.89k • 25 • 2
-
TigreGotico/model2vec-ATC-role-classification-bert-base-uwb-atcc
7.62M • Updated • 4 -
TigreGotico/model2vec-ATC-role-classification-potion-2M
1.95M • Updated • 7 -
TigreGotico/model2vec-ATC-role-classification-potion-4M
3.84M • Updated • 8 -
TigreGotico/model2vec-ATC-role-classification-potion-8M
7.62M • Updated • 5