NEWS.md

In bmsasilva/soundClass: Sound Classification Using Convolutional Neural Networks

soundClass 0.0.9

First public release.

soundClass 0.0.9.1

• Bug fix in overlap parameter of function spectro_calls(). Now accepts values of 0.5 and 0.75

• Include progress bar in console for function spectro_calls()

• Update bundled model code

• Update download link of example files

soundClass 0.0.9.2

• Bug fix in the window size parameter of function spectro_calls(). The y-axis values of the spectrograms generated are now correct

• Bug fix in metadata parameter of function auto_id(). The function now reads the file path correctly

• Added checkbox input for Butterworth filter in app_label()

• Minor updates in the documentation

soundClass 0.0.9.3

• Added function to plot training spectrograms

• Added parameter "frequency resolution" to function spectro_calls()

• Added functionality to apply a butterworth filter prior to recording processing. Respective parameter added to functions spectro_calls() and auto_id()

• Updated to remove deprecated function dbplyr::src_sqlite()

• The "Run model" GUI (in app_model()) now accepts butterworth filter and custom windows size for spliting the recording into chuncks

bmsasilva/soundClass documentation built on Jan. 10, 2023, 3:34 p.m.