This is a test set for German speech recognition, sourced from YouTube videos featuring Amazon Alexa.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
alexa
src
README.md
content.csv
test.py

README.md

german-sr-test

This is a test set for German speech recognition, sourced from YouTube videos featuring Amazon Alexa.

prefix

Each audio file's prefix specifies its source. Every source has different voice and sound quality.

prefix source
aos https://youtu.be/CXnL3rOVm-M
hkr https://youtu.be/8BGS6LUTMps
nwb https://youtu.be/vj-YUwY4rjc
vnx https://youtu.be/MajnL4j7DCk
fff https://youtu.be/5LJqZW4ZRfA

content

The content of each file can be found in alexa_content.csv.

scoring

In order to test a German speech recognition system, first generate a csv file in the same format as alexa_content.csv. Then run python test.py [filename].

requirements