Modelarts voice detection and text classification

Catalog

1. Voice detection and discrimination

1.1 Import data set import obs bucket

1.2 Create a sound classification project

1.3 model training

1.4 Deploy online

2. Text classification

2.1 from AIgallery Import datasets from

2.2 Create a dataset

2.3 Classify data sets

2.4 Create project

2.5 model training

2.6 Deploy online

---

1. Voice detection and discrimination

1.1 Import data set import obs bucket

1.2 Create a sound classification project

choice modelarts Sound classification items of

1.3 model training

After synchronizing the data source , Start labeling , After creating each label , You can mark ...

Here are the marked ：

Then you can train ... 

1.4 Deploy online

After training , You can deploy and go online , And tested ...

Start testing after deployment goes online ：

There are few datasets , So the effect is not very good ... But it was predicted

2. Text classification

2.1 from AIgallery Import datasets from

Download the takeout review data set

2.2 Create a dataset

Here you can see the data set that has not been marked ...

2.3 Classify data sets

By importing the dataset , And separate text and annotation , It can realize the effect of automatic learning and labeling ...

You can see here that it is half marked ... 

It's very sweet ...

2.4 Create project

  Select text classification items , Then create

2.5 model training

After publishing successfully ... Just wait until the training is finished

2.6 Deploy online

 

After training , It can be deployed online ...

 

You can see , Comments on takeout can still tell good from bad ...