Paul B Mahol
129978af6b
avfilter/af_arnndn: skip array extra entries by looking for new line
2021-02-10 14:15:42 +01:00
Paul B Mahol
199301a03a
avfilter/af_arnndn: make model opening errors more verbose
...
Also return proper error code.
2021-02-09 23:56:11 +01:00
Paul B Mahol
fb9c56bbd5
avfilter/af_arnndn: add support for commands
2021-02-05 00:43:11 +01:00
Anton Khirnov
c8c2dfbc37
lavu: move LOCAL_ALIGNED from internal.h to mem_internal.h
...
That is a more appropriate place for it.
2021-01-01 14:11:01 +01:00
Paul B Mahol
7025c91d07
avfilter/af_arnndn: add mix option
...
Mostly useful to listen to actual noise.
2020-12-04 23:57:01 +01:00
Paul B Mahol
5b1ccd748a
avfilter/af_arnndn: add timeline support
2020-11-25 13:41:52 +01:00
Paul B Mahol
939f4b35b8
avfilter/af_arnndn: reduce size of second_check[] table
2020-09-04 18:02:52 +02:00
Paul B Mahol
f7deae7215
avfilter/af_arnndn: use memcpy for copying in compute_rnn()
2020-08-27 20:39:33 +02:00
Paul B Mahol
ace1cc6163
avfilter/af_arnndn: use scalarproduct_float() in dct function
2020-08-27 20:39:33 +02:00
Paul B Mahol
999f5160c4
avfilter/af_arnndn: use RNN_COPY macro to copy
2020-08-21 22:46:07 +02:00
Andreas Rheinhardt
6aae322120
avfilter/af_arnndn: Remove superfluous qualifier
...
remove_doubling() returned a const float, but returning qualified types
is pointless anyway. This is all the more important given that GCC 4.2.1
(and maybe other compilers, too) seems to treat this as an error as several
FATE units that don't compile any more show.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
2019-10-17 12:30:24 +02:00
Paul B Mahol
fb7b222b91
avfilter/af_arnndn: make sure that vad_output nb_neurons is always 1
2019-10-17 11:02:02 +02:00
Paul B Mahol
b0bfa3699c
avfilter: add arnndn filter
2019-10-16 15:13:59 +02:00
