pollyana, branch main Voice input tooling for Linux. http://code.irenes.space/pollyana/atom?h=main 2025-09-07T02:09:55+00:00 a fancy new matching method defines commands with arbitrary suffixes 2025-09-07T02:09:55+00:00 Irene Knapp ireneista@irenes.space 2025-09-07T02:09:55+00:00 urn:sha1:a1d3db61140b397e30bd1e1d2142ef1467019eab Force-Push: yeah Change-Id: Ibe3a058173cded6355fadd413349cf9cc8823fdd the vanity commit (change kaylee to pollyana everywhere) 2025-09-06T23:22:50+00:00 Irene Knapp ireneista@irenes.space 2025-09-06T23:22:50+00:00 urn:sha1:63ed2e5a680b6e0e352721a8238e2918cb3e70ff except the copyright notices, those need more care and attention Force-Push: yes Change-Id: Ibddfb1b7a0edbb0d4adb6cfaf0ac4239537ade2f add a nix package for the now-forked pollyana 2025-09-06T23:04:36+00:00 Irene Knapp ireneista@irenes.space 2025-09-06T23:04:36+00:00 urn:sha1:18e38018cf4cc8f9689c453f5013426e60bf744a the contents of the patch against upstream kaylee are now applied to the pollyanna source; the patch is retained in case it's ever useful to build upstream Force-Push: yes Change-Id: I18df6c5b48ee1e767cff259bd323682c8dd693ff Merge kaylee into pollyana, yay 2025-09-06T22:46:23+00:00 Irene Knapp ireneista@irenes.space 2025-09-06T22:46:23+00:00 urn:sha1:7f96b7a15e3d688d8a679f03532b5b89101744ce (hope this works. never done it in git before) Force-Push: yeah... Change-Id: I71c8570d723808ccc8fbffea41c0722a61e1e1c0 move all the Kaylee code into a subdirectory 2025-09-06T22:42:27+00:00 Irene Knapp ireneista@irenes.space 2025-09-06T22:42:27+00:00 urn:sha1:cf6e693773101b5ac6aea7a4186e7e15ce4508d5 this is preparation for merging it into the Pollyana repo Force-Push: yes Change-Id: I155624ba39a0d212c999f38aaf2de2cd62b7aa49 initial package of kaylee 2025-09-06T22:25:26+00:00 Irene Knapp ireneista@irenes.space 2025-09-06T22:06:48+00:00 urn:sha1:e331e95e3f3607bd1707b60f61e209e4aeac05f6 Force-Push: working CI? that's unpossible Change-Id: I03dfbbcff92f2166616d26d4c3d556261a327d8b Fixed recognition of "forty" 2016-07-04T01:18:23+00:00 Clayton G. Hobbs clay@lakeserv.net 2016-07-04T01:18:23+00:00 urn:sha1:baeb0e1fa08fb6b21e4a22458de16b4948d1f2d0 I spelled it "fourty", which tricks the lmtool into making it much harder to recognize the word. Now it's spelled "forty", which makes it much easier to understand. Merge branch 'master' of https://github.com/Ratfink/kaylee 2016-06-29T17:16:31+00:00 Clayton G. Hobbs clay@lakeserv.net 2016-06-29T17:16:31+00:00 urn:sha1:26081985aeebfa15d6a68b801d0b82139a5080df Put all number-words on one line in the corpus 2016-06-29T17:14:58+00:00 Clayton G. Hobbs clay@lakeserv.net 2016-06-29T17:14:58+00:00 urn:sha1:cb4822025f2bd3e34ab39ca4a0cbef82a05d8f0d This makes pocketsphinx pick up way fewer false-positives of single number-words as recognised sentences. It doesn't seem to make any difference in anything else, but fewer false-positives is always nice. A bit of cleanup in the README 2016-05-31T19:33:32+00:00 Clayton G. Hobbs clay@lakeserv.net 2016-05-31T19:33:32+00:00 urn:sha1:a4c186b826c93765f0281ad28ff753a0d75fcc88 It actually does do the things claimed in the first sentence, so don't use the future tense. There's no need to tell the user the details of intermediate steps in generating the language. All the user needs to know is that it uses the network to update the language model when it starts.