clovaai/ClovaCall
ClovaCall dataset and Pytorch LAS baseline code (Interspeech 2020)
Listen-Attend-and-Tell (LAS) seq2seq architecture with convolutional feature extraction (2D convolution + bidirectional LSTM encoder) and attention-based decoder for Korean goal-oriented dialog ASR. The dataset contains 112K utterance-transcript pairs from contact center interactions in restaurant reservation domain, with both raw and silence-eliminated versions totaling ~125 hours of multi-speaker Korean speech. Code supports PyTorch-based training with data augmentation techniques (noise augmentation, SpecAugment) and evaluation metrics using character error rate (CER).
223 stars. No commits in the last 6 months.
Stars
223
Forks
55
Language
Python
License
MIT
Category
Last pushed
Apr 05, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/clovaai/ClovaCall"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
TensorSpeech/TensorFlowASR
:zap: TensorFlowASR: Almost State-of-the-art Automatic Speech Recognition in Tensorflow 2....
xinjli/allosaurus
Allosaurus is a pretrained universal phone recognizer for more than 2000 languages
dangvansam/viet-asr
VietASR - Vietnamese Automatic Speech Recognition
wenet-e2e/wenet
Production First and Production Ready End-to-End Speech Recognition Toolkit
srvk/eesen
The official repository of the Eesen project