I realize that this might not be the most well-formulated question, but I figured I'd throw it out there as an open-ended discussion that could turn up some interesting input.

I'm relatively new to speech processing, and came across the statement "invariance to transformation is important in many speech engineering tasks" in a paper I'm reading.

I was wondering if anyone could provide any specific examples that could illustrate what exactly this means. The word transformation to me is a bit ambiguous, maybe because I'm relatively new to this particular field, but I was hoping that this statement is a little more obvious to someone else out there who is more familiar with the field.

  • 1
    $\begingroup$ Maybe you could cite the paper so we could see it in context (if the article is freely available). $\endgroup$ – Jason R Jul 18 '13 at 18:14
  • $\begingroup$ Actually I think I may have come up with an answer. I think all the author is trying to say is that if some signal attribute varies across a single segmentation class, then it may not be a useful attribute to segment using. $\endgroup$ – wheresmycookie Jul 18 '13 at 18:17
  • $\begingroup$ The paper, for those curious, is "A Study on unsupervised phoneme segmentation and its application to automatic evaluation of shadowed utterances". The statement I was referring to is at the end of the second page. $\endgroup$ – wheresmycookie Jul 18 '13 at 18:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.