Ask HN: Self-hosted text-to-speech (TTS) solutions that can parse PDFs?
Anyone know of any solutions that perform TTS on PDFs and the such?
Basically something like https://speechify.com but open sourced. The motivation, if not already apparent, is to self-host and not pay for something that maybe easily bubblegum + taped together....maybe.
2 comments
[ 3.3 ms ] story [ 8.4 ms ] thread1) Extract text from PDF 2) TTS on extracted text
Not hard to glue together. You can check the eco system in your favorite programming language
I found the combination of the NVDA screen reader and Adobe Acrobat works well from an accessibility standpoint, particularly that combo does the right thing for the tables we tested it on. I tried some off-brand PDF readers and got worse results.