Add sorting options for non-time-constrained WERs #49
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
boeddeker
reviewed
Jan 25, 2024
meeteval/io/seglst.py
Outdated
| 'Invalid JSON format for SegLST: Expected a list of segments, ' | ||
| 'but found a dict.' | ||
| ) | ||
| if loaded and not isinstance(loaded[0], dict): |
There was a problem hiding this comment.
The first entry in a json is sometimes special. We should also check the last entry (last entry avoids more ifs):
if loaded and not (isinstance(loaded[0], dict) and isinstance(loaded[-1], dict))
There was a problem hiding this comment.
Done, also checked for "session_id" key.
boeddeker
approved these changes
Jan 25, 2024
thequilo
added a commit
that referenced
this pull request
Jan 25, 2024
* Add sort options to non time-constrained WERs * Fix --regex option in _merge * Convert boolean args (True, true, False, false) to bool * Add burn tests for sort options * Improve error message for seglst loading * Remove debug help text * Improve seglst format check
thequilo
added a commit
that referenced
this pull request
Jan 25, 2024
* Add visualization * Add visuals - Options to hide legend and details - Add markers to the utterance begin/end to indicate direction - Add hint for empty segments * Fix minimap selection * Drag details view to scroll * Add Minimap x-axis label * Fix legend * Display some information about the example * Add menu * Throttle redraw of details plot * Add more options for minimap display * Fix match lines * Move axis into own classes and improve visuals * Add message when no words are displayed in details plot * Update colormap for readability * Display info about selected segment * Mark selected segment in details plot * Fix empty label * Zoom centered around mouse * Move plot label style to css * Draw boxes around plots * Set font to Arial * Add basic support for touch devices * Draw background in details plot * Disable warnings in visualization * Update details view * Remove debug alerts * Update visualization for tidy data * Add WER decomposition tooltip * Remove brush when fully zoomed out * Add range selector with text input * Simplify callback code for fetching audio * Add search bar * Add alignment_transform function - Adds ability to ignore words in the matching but display them in the visualization * Add tooltip table for segment details * Fix zoom in details view when brush is removed in minimap * Use Tidy representation for visualization and add (rudimentary) markers * (WIP) Tidy -> SegLST * A few fixes - json -> simplejson - centered and growing tooltips - drawing details * Fix CSS so that fontawesome icons get displayed * Cleanup visualization code * Improve tooltip positioning * Cleanup visualization * More cleanup * Move tooltip fixing code into a function * Add audio players when an audio path is found in a segment * add a file server that normalizes the audio (usefull for far field data) * viz: Add copy button and try fileserver before acessing local files - Fileserver works in jupyter, but requires a process in the background - The fileserver can normalize the audio files - Local files work only for standalone viz * viz: ogg is more robust * viz: Fix infinity loop and remove play button it it doesn't work. * viz: Update audio help text * viz: file server: - Always answer with HTTPUnauthorized if the audio cannot be loaded - Use soundfile to remove sox dependency - Use ogg, because wav produced artefacts in some chrome and firefox versions * add aiohttp to cli dependencies * fix test dependency * Add sorting options for non-time-constrained WERs (#49) * Add sort options to non time-constrained WERs * Fix --regex option in _merge * Convert boolean args (True, true, False, false) to bool * Add burn tests for sort options * Improve error message for seglst loading * Remove debug help text * Improve seglst format check * Fix seglst format check * add soundfile as cli dependency * support http files, verbose keyerrors and use operator.attrgetter * fix load for http * Also convert int when parse_float != float * Fix orthography field name * Add viz to readme * SegLST: type annotations * fix py37 --------- Co-authored-by: Christoph Boeddeker <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.