This is a draft dataset of prosodic minimal pairs. It includes two stress doublets, the noun and verb uses of object and of present.Remaining words are from superb ks.
object10 and present10 tokens lexically have left stress. object01 and present01 tokens lexically have right stress.
There are separate validation splits for each of object10, present10, object01 and present01. In each of these splits, all the tokens
have the same segmental content and lexical prosody. Test splits are not… See the full description on the dataset page:
https://huggingface.co/datasets/MatsRooth/object_v.
