Skip to content

[ICASSP2023] Source code, model links and open test sets for paper SeACo-Paraformer.

Notifications You must be signed in to change notification settings

R1ckShi/SeACo-Paraformer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 

Repository files navigation

SeACo-Paraformer

Introduction

Semantic Augmented Contextual-Paraformer (SeACo-Paraformer) is a non-autoregressive E2E ASR system with flexible and effective hotword customization ability which follows the main idea of CLAS and ColDec. This repo is built for showing (1) detailed experiment results; (2) source codes; (3) open model links; (4) open hotword customization test sets based on Aishell-1 as discussed in our paper.

Paper

SeACo-Paraformer: A Non-Autoregressive ASR System with Flexible and Effective Hotword Customization Ability

Source Codes

The model proposed and compared in the paper are implemented with FunASR , which is a open source ASR toolkit.

  1. Paraformer Model : source code
  2. Paraformer-CLAS Model : source code
  3. SeACo-Paraformer Model : source code

Model Links

Modelscope is a open platform for sharing models, AI spaces and datasets. We open our industrial models discussed in the paper through Modelscope:

  1. Paraformer Model: model link
  2. Paraformer-CLAS Model: model link
  3. SeACo-Paraformer Model: will be released once paper is accepted: model link

How to Reproduce the Results

With FunASR toolkit, you can use many ASR models to recognize, which include different language ASR models, different model architectures like SeACo-Paraformer

The following code conducts ASR with hotword customization:

from funasr import AutoModel

model = AutoModel(model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
                  model_revision="v2.0.4",
                  device="cuda:0"
                  )

res = model.generate(input="YOUR_PATH/aishell1_hotword_dev.scp",
                     hotword='./data/dev/hotword.txt',
                     batch_size_s=300,
                    )
fout1 = open("dev.output", 'w')
for resi in res:
    fout1.write("{}\t{}\n".format(resi['key'], resi['text']))

res = model.generate(input="YOUR_PATH/aishell1_hotword_test.scp",
                     hotword='./data/test/hotword.txt',
                     batch_size_s=300,
                    )
fout2 = open("test.output", 'w')
for resi in res:
    fout2.write("{}\t{}\n".format(resi['key'], resi['text']))

Open Hotword Customization Test Sets

Previous hotword customization related works have seen performance reported on internal test sets, we hope to provide a common testbed for testing the customization ability of models.

Based on a famous Mandarin speech dataset - Aishell-1, we filtered two subsets from test and dev set and prepare hotword lists for them. Find them in data/test and data/dev, you may download the entire Aishell-1 and filter out the subset Test-Aishell1-NE and Dev-Aishell1-NE with data/test/uttid and data/dev/uttid respectively. And we prepare the entire hotword list hotword.txt and R1-hotword list r1-hotword.txt for them (R1-hotword stands for hotwords whose recall rate on a general ASR model recognition results is lower than 40%)

#utt #hotwords #R1-hotwords
Test-Aishell1-NE 808 400 226
Dev-Aishell1-NE 1334 600 371

Experiment Results

In the Paper

Result1: Comparing the models

Result2: Ablation study over bias decoder calculation

Result3: Performance comparision with larger incoming hotword list

Result4: Attention score matrix analysis

About

[ICASSP2023] Source code, model links and open test sets for paper SeACo-Paraformer.

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published