ibus-engines.stt: init at unstable-2017-12-30

[jtojnar]

Description of changes

Accidentally found a Hackaday article mentioning ibus-deepspeech in the bookmarks of my RSS feed reader so decided to try it out. Updated stt while at it.

I managed to make the example from gst-deepspeech work with upstream audio samples but not with my microphone – perhaps. The ibus engine unfortunately does not seem to do anything for me either.

Edit: It must be some issue with the gstreamer pipeline since when I record my voice and play it, it works :

$ nix-shell -I nixpkgs=$HOME/Projects/nixpkgs -p gst_all_1.gstreamer -p gst-deepspeech --run 'gst-launch-1.0 --messages filesrc location=$PWD/fox.flac ! decodebin ! audioconvert ! audiorate ! audioresample ! deepspeech silence-length=20 ! fakesink | grep -E deepspeech.+text'
TensorFlow: v2.8.0-8-g06c8fea58fd
 Coqui STT: v1.3.0-0-g148fa743
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.
Got message #169 from element "deepspeech0" (element): deepspeech, timestamp=(guint64)18446744073709551615, stream-time=(guint64)18446744073709551615, running-time=(guint64)18446744073709551615, intermediate=(boolean)false, text=(string)"to\ be\ bound\ to\ jumped\ over\ the\ lazy\ dog";

Whereas the pipeline from the ibus engine is way too off:

$ nix-shell -I nixpkgs=$HOME/Projects/nixpkgs -p gst_all_1.gstreamer -p gst-deepspeech --run 'gst-launch-1.0 --messages pulsesrc ! audioconvert ! audiorate ! audioresample ! deepspeech silence-length=20 ! fakesink | grep -E deepspeech.+text'
TensorFlow: v2.8.0-8-g06c8fea58fd
 Coqui STT: v1.3.0-0-g148fa743
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.
Got message #62 from element "deepspeech0" (element): deepspeech, timestamp=(guint64)18446744073709551615, stream-time=(guint64)18446744073709551615, running-time=(guint64)18446744073709551615, intermediate=(boolean)true, text=(string)"for\ the\ ";
Got message #63 from element "deepspeech0" (element): deepspeech, timestamp=(guint64)18446744073709551615, stream-time=(guint64)18446744073709551615, running-time=(guint64)18446744073709551615, intermediate=(boolean)true, text=(string)"for\ the\ wise";
Got message #64 from element "deepspeech0" (element): deepspeech, timestamp=(guint64)18446744073709551615, stream-time=(guint64)18446744073709551615, running-time=(guint64)18446744073709551615, intermediate=(boolean)true, text=(string)"for\ the\ wise";
Got message #65 from element "deepspeech0" (element): deepspeech, timestamp=(guint64)18446744073709551615, stream-time=(guint64)18446744073709551615, running-time=(guint64)18446744073709551615, intermediate=(boolean)true, text=(string)"for\ the\ wise";
…

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 22.05 Release Notes (or backporting 21.11 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[jtojnar]

There is also some work to make stt plug-in part of gstreamer but not ready, AFACT: Elleo/gst-deepspeech#19 (comment) https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1749

[nixos-discourse]

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/speech-to-text-virtual-keyboard-in-nixos/22684/2