Pinned Repositories
MASR
Pytorch实现的流式与非流式的自动语音识别框架，同时兼容在线和离线识别，目前支持Conformer、Squeezeformer、DeepSpeech2模型，支持多种数据增强方法。
pytorch_MLP_for_ASR
This code implements a basic MLP for speech recognition. The MLP is trained with pytorch, while feature extraction, alignments, and decoding are performed with Kaldi. The current implementation supports dropout and batch normalization. An example for phoneme recognition using the standard TIMIT dataset is provided.
conformer
PyTorch implementation of "Conformer: Convolution-augmented Transformer for Speech Recognition" (INTERSPEECH 2020)
dlbeginners
神经网络与深度学习--上海财经大学涂老师
end2end-asr-pytorch
Low Rank Transformer ASR
hypermixing
PyTorch implementation for HyperMixing, a linear-time token-mixing technique used in HyperMixer architecture
MASR111
PPASR
基于PaddlePaddle实现端到端中文语音识别，从入门到实战，超简单的入门案例，超实用的企业项目。支持当前最流行的DeepSpeech2、Conformer、Squeezeformer模型
SoftPool
[ICCV 2021] Code for approximated exponential maximum pooling
MASR
Pytorch实现的流式与非流式的自动语音识别框架，同时兼容在线和离线识别，目前支持Conformer、Squeezeformer、DeepSpeech2模型，支持多种数据增强方法。
CSLJingyu's Repositories
CSLJingyu/MASR111
CSLJingyu/hypermixing
PyTorch implementation for HyperMixing, a linear-time token-mixing technique used in HyperMixer architecture
CSLJingyu/MASR
Pytorch实现的流式与非流式的自动语音识别框架，同时兼容在线和离线识别，目前支持Conformer、Squeezeformer、DeepSpeech2模型，支持多种数据增强方法。
CSLJingyu/conformer
PyTorch implementation of "Conformer: Convolution-augmented Transformer for Speech Recognition" (INTERSPEECH 2020)
CSLJingyu/PPASR
基于PaddlePaddle实现端到端中文语音识别，从入门到实战，超简单的入门案例，超实用的企业项目。支持当前最流行的DeepSpeech2、Conformer、Squeezeformer模型
CSLJingyu/SoftPool
[ICCV 2021] Code for approximated exponential maximum pooling
CSLJingyu/dlbeginners
神经网络与深度学习--上海财经大学涂老师
CSLJingyu/end2end-asr-pytorch
Low Rank Transformer ASR
CSLJingyu/pytorch_MLP_for_ASR
This code implements a basic MLP for speech recognition. The MLP is trained with pytorch, while feature extraction, alignments, and decoding are performed with Kaldi. The current implementation supports dropout and batch normalization. An example for phoneme recognition using the standard TIMIT dataset is provided.