Issue with Nasal Vowels

[VikasBalani02]

While using the epitran library for generating the IPA for Hindi, I found that the library doesn't give the expected output.
For example: the input "ऑस्ट्रेलिया" to the transliterate function gives the output "ऑsʈrelijaː" the expected output is "ɒsʈrelijaː".

[ruohoruotsi]

I've also seen inconsistent results for Yorùbá nasalized (esp. tonal) vowels. I'll be submitting a pull-request to address this soon for Yorùbá. I can't help with Hindi, but will be curious to learn if/how this gets resolved.