EVA

 Week 11:  We built an extension that can compose the subject and body of an email through voice. The user will have to use " line break " command to differentiate between subject and body  The colors were optimized to comfort the users with color blindness as well. As per our research most color blind users have a hard time recognizing between red, blue and green. So we decided to use shades of green and blue. Presently the UI consists of a gradient background colors of green and blue which should be comforting to all users.  All of these are presently baked into the popup UI of our extension and in the next week we will be looking into the webpage operations through voice.   Week 12: This week was one hell of a ride. We could give a voice output of the details of a specific email. All of these are achieved through voice commands which are required to be specified by the user. We have added these following commands : "read one", "read second", "read ...

 Week 10(14/06/2021) We integrated the UI with the part that was running on localhost which was retrieving emails and deployed it on replit. We also added a few voice commands such that total mail count and unread mail count is given as a voice output. We did research on colors such that someone with color blindness would also be able to use our interface easily and differentiate elements   So far we have a UI which can be comforting to color blind users, retrieval of 10 emails from the user's inbox and a voice output of unread emails count of the user.  Challenges: When a voice command is given the output is repeated twice and we are finding it a bit challenging while getting voice outputs of retrieved emails.

 Week 09(07/06/21): We found some challenges in regards to calling multiple Gmail APIs directly in an extension, so for now we have resorted to using a localhost server to call these APIs and run all the important functionality of our project on it till we figure out how we can resort to calling multiple APIs in the extension. We also made a mock-up UI of our prototype which has a simple layout. One of the main intention of this UI is to make the extension screen-reader friendly as well. We were able to create a compose window using Gmail API from which mails can be sent, further all of it will be revamped with the UI and will see how we can go further from there We will be testing out a few test cases in the coming weeks like: Checking how to access user created folder in gmail. To find a solution as to where the contents of the mail can be stored if the user loses internet connection before sending the mail.  

 Week 8: (31/05/2021) During our 1st review to the department we had specified that the "@" and "at" recognition was a bit challenging, so we modified the speech input method in such a way that if the user says the username part of the email address and upon ending his speech input the extension would append "@gmail.com" to the text. But what about other email clients? We are looking into how to interface with the user if he wants to change the client name of the email address. We also might change this whole method so be tuned in for that.  We also looked into gmail.js, A package which can used as an alternative to Gmail APIs for all CRUD operations, but we found few drawbacks such as It mainly requires the gmail window to be open for the functioning of the extension Slows down the load time of gmail window If you have too many functions specified you may get your gmail account suspended ( speaking from experience, do not recommend that, the account did g...

  Week 6: Built an extension that can call Gmail APIs to get information like id, unread emails count etc We logged those information on the console Found out that each email can have a different mime type like text/HTML and text/plain Implemented Artyom.js' command functionality into our extension    Week 7:   Used Artyom.js' command functions to switch between a subject box and a compose box in our popup We also could print the unread emails count on our extension's popup display Made our extension to update it's unread emails count for every 1 minute Retrieved text/plain mime type of an email   Next steps:  Iterate around with the commands function to see what suits best for our email operations   Commands interfering with the content in subject and body   Reading  attachments, Formatted contents, dealing with email threads. Noisy environment might pose a challenge for the speech synthesis of the user’s audio input.  

 Fifth Week Progress(10/05/2021): Added a textbox which can be used to store in all the textual content which the user wants to mail. Used Speech Recognition to recognize all the contents the user would speak out and convert it into a textual format. Used artyom.js throughout for speech synthesis  Started testing out the Commands feature available in the artyom library. Explored the speech synthesis package p5.speech.js to build a basic webpage which speaks out hardcoded text Made a few changes in the basic extension which used chrome.tts.speech to speak out hardcoded text   In the next week:   We will be integrating all of this an extension for testing purposes. We also will be testing out other stt and tts libraries to use in our project.