225 commits

This branch

This branch

All branches

Author	SHA1	Message	Date
dscripka	c8419358a6	More work on documentation	2022-11-15 21:19:22 -05:00
dscripka	a2c03f828c	Significant work to README	2022-10-30 21:47:24 -04:00
dscripka	f14c88a0a0	Update github actions runner config	2022-10-30 17:16:02 -04:00
dscripka	137dead059	Merge branch 'main' of https://github.com/dscripka/openWakeWord into main	2022-10-30 17:11:57 -04:00
dscripka	f45a8b4e36	Fix issues with flake8 version	2022-10-30 17:11:47 -04:00
dscripka	52aada7485	Setup Github Actions for CI/CD	2022-10-30 17:02:49 -04:00
dscripka	36a5c6e9ad	Fixed typos in requirements	2022-10-30 17:00:43 -04:00
dscripka	bc84c2edba	remove old test model	2022-10-29 15:05:57 -04:00
dscripka	89230ca697	All tests (including flake8 and mypy) testing, but coverage is still low	2022-10-29 14:44:05 -04:00
dscripka	2ccb903d61	Small changes/bug fixes, starting refactor in preparation for first release	2022-10-27 21:13:30 -04:00
dscripka	cb7a06ac67	Added ability to apply reverberation when using the mix_clips utility function	2022-10-12 22:40:05 -04:00
dscripka	465c339fda	Notebook updates, model update for TTA, though might remove that in the future if simple filtering works better	2022-10-11 23:16:04 -04:00
dscripka	e97e704287	Fixed bugs in a few functions, added some more examples notebooks	2022-10-10 23:41:11 -04:00
dscripka	c1466374ba	Changes some filenames, started to work on docs, added a function	2022-10-09 23:45:09 -04:00
dscripka	623e86f505	Added tta (test time augmentation) method to reduce false positive predictions only when an initial positive prediction is made	2022-10-08 10:52:50 -04:00
dscripka	cd919c73cc	Changed how mmap batching function computes sizing, few bugfixes	2022-10-08 01:05:42 -04:00
dscripka	2f3e221ea5	Few bugfixes, added function to predict on an audio file in simulated streaming mode	2022-10-05 23:18:35 -04:00
dscripka	478fb0f5ae	Added example script for real-time microphone inference	2022-09-28 22:50:34 -04:00
dscripka	37e37722c4	Modifications to data preparation functions, better duration estimates for non-WAV audio files	2022-09-27 22:54:34 -04:00
dscripka	ea82bc4ae3	Moved some functions to new files, standardized audio loading to always ensure that audio data is always loaded as 32-bit float, -1 to 1 range	2022-09-26 23:10:35 -04:00
dscripka	b52c33043e	Added more example notebooks, small bug fixes and changes	2022-09-25 23:01:02 -04:00
dscripka	ca66ca9c4c	Added several new utility methods for batch processing, as well as first example notebook	2022-09-21 22:01:28 -04:00
dscripka	a8949f3b48	Set fix input frame size to simplify code, added benchmark script to estimate inference efficiency	2022-06-21 08:32:32 -04:00
dscripka	8d89349179	Initial commit, basic model implemented and first test passing	2022-05-30 21:03:13 -04:00
dscripka	3d0b69ee46	Initial commit	2022-05-28 15:09:35 -04:00