paloukari/OrcaDetector

Issues

• Download the data

  #61 opened 3 years ago by quentin-hamard
  1

• Docker container for the TX2

  #30 opened 6 years ago by mwinton
  1

• live feed proxy

  #23 opened 5 years ago by mwinton
  1

• Manually annotate audio samples to find the important segments

  #13 opened 5 years ago by mwinton
  2

• Create central CLI entry point for the various scripts

  #19 opened 5 years ago by mwinton
  1

• Investigate if there is any value to keep the trailing segment of the audio files

  #28 opened 5 years ago by paloukari
  1

• Capture noise data

  #33 opened 5 years ago by mwinton
  1

• Remove vggish smoke test, add orca smoke test

  #17 opened 5 years ago by mwinton
  1

• Decide which species to explicitly classify

  #5 opened 5 years ago by mwinton
  1

• Record loss and accuracy after each training epoch; generate plots

  #2 opened 5 years ago by mwinton
  1

• Set up MLFlow logging server for recording results from experimental runs

  #3 opened 5 years ago by mwinton
  1

• Add argparser support to specify a model name

  #7 opened 5 years ago by mwinton
  1

• Add support for running test

  #9 opened 5 years ago by mwinton
  1

• add Batch Normalization before feeding audio into model

  #21 opened 5 years ago by mwinton
  0

• Add a brief /vggish/README.md

  #1 opened 6 years ago by mwinton
  1

• Add --verbose CLI flag to trigger debugging output

  #8 opened 6 years ago by mwinton
  1

• Additional EDA plots

  #27 opened 6 years ago by mwinton
  0

• Dockerize the project

  #10 opened 6 years ago by mwinton
  0

• Need to improve performance of the Keras generators

  #15 opened 6 years ago by mwinton
  0

• Update to support shorter (2 sec?) audio clips

  #14 opened 6 years ago by mwinton
  0

• Investigate if/how to reduce the used classes

  #29 opened 6 years ago by paloukari
  0

• 6/25 project intro for class

  #24 opened 6 years ago by mwinton
  0

• notebook for exploring dataset

  #22 opened 6 years ago by mwinton
  1

• Use https://github.com/asottile/future-fstrings to tidy up our string formatting

  #20 opened 6 years ago by mwinton
  2

• Remove links to download the tar archive containing the audio files

  #18 opened 6 years ago by mwinton
  0

• track down and give credit to original author of the web scraper

  #16 opened 6 years ago by mwinton
  0

• Set up train/val/test split of our downloaded data

  #4 opened 6 years ago by mwinton
  0

• Create the vggish Keras model

  #6 opened 6 years ago by mwinton
  1