Joll59/d-ser-t

Issues

• Note the proper bit rate for audio files

  #90 opened 5 years ago
  0

• Add KER metric functionality

  #89 opened 5 years ago
  0

• Remove work for unhandledSTTOutput

  #88 opened 5 years ago
  0

• Blank response when streaming files through but when singular file is sent transcription is received

  #83 opened 5 years ago
  1

• XmlWriterService and TranscriptionFileService not showing up in code coverage report

  #81 opened 5 years ago
  0

• Create an interface to work with any number of file types.

  #80 opened 5 years ago
  0

• [Bug] Concurrent calls exceeding 5, results become inconsistent

  #71 opened 5 years ago
  0

• TranscriptionAnalyzerCI does not factor in the usage of the same word twice in an utterance when running cleanActualTranscription

  #65 opened 5 years ago
  0

• eliminate log statements in package.

  #47 opened 5 years ago
  0

• Document global install of d-ser-t-cli

  #46 opened 5 years ago
  1

• Cleaning actual transcriptions produces the same output as cleaning expected transcriptions.

  #40 opened 5 years ago
  2

• Consider making TranscriptAnalyzer a static class

  #39 opened 5 years ago
  1

• Move literals into a config file

  #38 opened 5 years ago
  0

• Bug : setting the wrong concurrency number crashes the tool

  #35 opened 5 years ago
  1

• Match audio files with transciption than the vice versa

  #33 opened 5 years ago
  0

• validate audio file extension

  #31 opened 5 years ago
  0

• Any work done more than once should be a global/instance variable

  #29 opened 5 years ago
  1

• [Proposal] Standardize Commit Messages.

  #26 opened 5 years ago
  0

• need to handle non-windows file paths

  #23 opened 5 years ago
  2

• JSON results should include expected Transcription

  #22 opened 5 years ago
  1

• refactor TranscriptionService Class for Testability

  #21 opened 5 years ago
  0

• Consider the need for handleResponse function

  #20 opened 5 years ago
  1

• Force the node version to 12.4

  #19 opened 5 years ago
  0

• Convert d-ser-t to a Mocha test

  #18 opened 5 years ago
  2

• Update package.json to allow for publishing the tool with proper versioning

  #17 opened 5 years ago
  1

• Word error rate should have 3 significant figures

  #16 opened 5 years ago
  0

• Update ReadMe to latest bits

  #15 opened 5 years ago
  0

• Consider calculating SER based only on significant words

  #14 opened 5 years ago
  0

• Refactor textfile parsing to be unit testable

  #13 opened 5 years ago
  1

• Validate transcription.txt early so that we fail fast

  #12 opened 5 years ago
  1

• Resolve typing issue in main.start()

  #11 opened 5 years ago
  2

• Verify CLI arguments

  #10 opened 5 years ago
  0

• Consider whether or not we should close the recognizer

  #9 opened 5 years ago
  1

• Transcriptions with false negatives should not be graded

  #8 opened 5 years ago
  0

• Analyze actual transcription and save output that is currently unhandled

  #7 opened 5 years ago
  1

• successRate should be in the range of 0-100

  #6 opened 5 years ago
  1

• successRate and avgUtteranceErrorRate should be measured on the same scale

  #5 opened 5 years ago
  0

• 2 incomplete results from a single .wav file

  #4 opened 5 years ago
  4