Test lists for the synthesized speech-recognition test "Oldenburg Phrases" (OLPHRA)

New speech material consisting of phrases was synthesized using a commercial text-to-speech system (Acapela Cloud Service). The phrase corpus is described in [1] and the phrases (audio files) are free avaible at Zenodo [2]. The phrases were combined into 1000 phonemically and perceptually equivalent test lists consisting of 20 phrases each. More details can be found in [3]. Files:OLPHRA_PhrasesPerTestList.csv The table shows the final composition of the OLPHRA test lists. Each row represents one test list. The first column indicates the test list number, followed by the phrases assigned to that list. The phrase number is equivalent to the file name of the corresponding audio file.