hauptdigital / deepspeech-notes Goto Github PK

User story

Users need to be informed about the purpose of the project, used technologies (+ related acknowledgements) and the author.

Desired

Additional page that shows

• purpose of the project
• author of the project with social media / GitHub links
• list of technology stack
• acknowledgements for inspiration and used resources / modules

React components are needed that display text chunks from transcribed audio

To work as intended, the application needs to use a set of dependencies. As a package manager, npm or yarn might be used to install these packages / modules:

The machine that will be used to develop the application needs to have:

• npm
• Node.js
• MongoDB
• sox
• Python

First:

• initialize yarn or npm

• Dev dependencies:

  • Code Quality (these are needed for front- and backend):
    • ESlint
    • Prettier
      compatibility with older browsers.

• Production dependencies:

  • Frontend
    • React
    • Storybook
    • emotion (styled components)
    • Speech-to-text functionality
      • socket.io
      • node-vad
  • Backend
    • Express.js

The repository should have a project introduction in the README.md that contains:

• Brief introduction of the project

When the project is ready for the first release, the README.md needs to be completed with all needed information. This will be handled in another issue.

To start development, a design is needed that provides all necessary resources (images, styles, colors) to develop all pages / components.

These are the steps to complete a full design template:

• Create mood board
• Find color scheme
• Create design for each page and component named in [todo] in Adobe XD
• Run a design review with code buddies
• Finalize design with approval from code buddies

User story

Users need to be able to edit text chunks in a note with their device keyboard as the speech-to-text feature "only" has a reported 95% accuracy. Users should also be able to delete a text chunk.

Desired functionality

• When a text chunk is focused, a cursor appears to edit the text chunk
• When a text chunk is edited, every change in a text chunk is saved directly

A React component is needed, that will display a microphone icon that shows the record state (listening / inactive).

User story

Users of the app should be able to create text notes when they speak into their browser microphone. To create a note, they should be able to activate / deactivate the microphone of their device to control, when the app is listening for voice input that should be created as a note.

Desired functionality

• When the microphone button is pushed, it is toggled active / inactive
• When the recorder is active, it should listen for input and display the transcribed text chunks as deepspeech returns them.
• Text chunks should be displayed as they come in from the speech recognition engine
• When the recorder is active, it should display an audio wave form to show noise activity detection

README.md needs to be updated with the information:

• Description and goal of the project
• Installation guide if someone wants to use deepspeech-deepsea
• Used technologies and acknowledgements
• Information about the author

User story

Users need to be able to completely remove notes.

Desired functionality

• When editing a note, a button should enable the user do delete a note.

Add React components and functionality related to the navigation. Menu should slide in from the right.

User story

Users need to be able to see previously created notes in an organized list. By looking at the note, they should be able to see a teaser preview of the content of the note. A click on the note should direct the user to the specific note.

Desired functionality

• In case of no previously created notes, display a message that informs, that there are no notes and the user should go ahead and create the first note.
• Display a list view of notes that contains cards showing a note title (if available) and a teaser text of the note.
• A click on a note in the list will open the full note for further editing.

A new component needs to be created that displays an audio waveform.

User story

Users should be able to see a statistic of previously created notes.

Desired functionality

Show users a statistic that contains:

• Number of notes
• Time of speech input that has been transcribed so far.
• Number of words that have been created by speech-to-text transcription so far.

Before the design can continue (#2 ), there is a need to define all pages of the application, so we know what pages need to be designed. In addition, for each page, there is a need to list all functionalities (i.e. user stories) to also define different parts on the pages.

• Define page structure and navigation
• Define functionality for each page (What tasks can be achieved by the user? )