Quill Digital. What's new in this release?
AWS Textract has arrived!
Textract is now available as an OCR option for the Quill Online Portal, application and our API.
Changes to OCR engines
We now include Google Cloud Vision, Microsoft Vision, AWS Textract and our own Quill Light OCR engine. These can be chosen in the settings if using the application, the settings on the Quill Portal or specified as 0,1,2,3 if using the Digitise web service call in the API.
Grayscaling and De-skewing
A grayscale option has been added to enhance any rough scans that are sent. This will add time to processing, but will de-skew and grayscale scanned documents before OCR to give a better read for text.
Audio files(wav,MP3) change of processor
Audio is now transcribed using a different processor for a better transcription. While there is no change to licensing for this, it will be faster and more accurate. Transcriptions will also include an accuracy rating.
Natural Language Processing processor
Included in the release is a powerful NLP processor. Simply pick the token and Quill will return the tokenised document. Valid tokens are:
New translation processor
Translation of Audio and Documents can now be run via the application, Portal and API. Pick from a list of predefined languages that Quill will translate.
Field Extraction will now save and run any custom regex that are given to Quill Digital. It will return these as a standard field extraction in the Application, Portal and API.
Addition of ability to amend and cancel billing subscriptions from the online portal.
Quill Digital. Active Client Change Requests
Below are current active requests from Quill Digital Clients that our developers are working on for the next release.
Addition of ability to choose a language code for Audio. Currently Quill Digital will only recognise EN-GB as Audio, and while it will run other languages it will not return an accurate transcription.
Addition of XML/JSON returns from API. Currently Quill Digital will only return raw data, unformatted for all requests.
Addition of abilty to request whether or not an accuracy rating is returned in Audio transcription. Currently Quill Digital returns audio transcripts with an accuracy rating per line.
Addition of ability to send more audio file types. Currently Quill Digital will only process .WAV and .MP3 files.
Quill Digital. Active and Known Bugs
Below are current active bugs in Quill Digital that our developers are working
Occasionally audio transcripts can miss words and the end of a sentence.
Occasionally translation returns no data on the online portal.
Large documents do not translate when using the Google OCR method.