Increment 1 - a190884810/CS5551 GitHub Wiki

CSEE 5540 0001 Special Topics Spring 2019



PROJECT INCREMENT -1

Team members:

  • Xie, Tiancheng class ID: 25
  • Doss, Corey Jason class ID: 8
  • Sun, Chen class ID: 23
  • Dhabbah, Khalid class ID: 6

Increment Content

  1. Detail Design of Features (using tools)
  2. Testing
  3. Implementation
  4. Existing Design/Services/REST API
  5. Deployment
  6. Project Management
  7. Bibliography



Detail Design of Features

  • For the Increment 1, considering all our team has no great Web/mobile application development experience. We would like to apply all that knowledge what we have learned in the class for part of our project work. In terms of, working on a single module of Model as a service, we would prefer building the framework of the whole application rather than thinking of how to do the connection between the backend and the frontend.

  • From the past few weeks, we have learned the following things

    • For the development platform, we decided to use VScode rather than Webstorm.
    • Some Front End development using HTML5, CSS3, JS, BootStrap, Angular 7, AJAX
    • For Single Page Application development we will use WebServices Mashup (Angular 7)
    • Ionic knowledge for mobile application development.

  • Architecture:


  • User story
  • As a user, I would like to achieve the image processing with the existed dataset. I will search image model, upload the image data and get the result from the model page.

Unit Testing

  • Considering that is the first phase of the project. We do not have much more work to perform unit testing. But for reinforcing what we have learned in the class, we would like to try some testing on those work based on the assignment which could be an important source for the project.

  • For example, a snip of unit testing in lab4 assignment:



Implementation

  • When our team was doing research on how to launch a model onto web/ mobile, we focused on tensorflow.js and tensorflow lite. The former one is a Javascript library that using node.js to deploy trained model in the web. Relatively, tensorflow lite is the one can help us launch model on the app. The limitation is, it is hard to figure out the theory on those deep-learning work. So far, we decide to retrieve some of those models source code as reference.

  • In addition, we would like to build more interfaces with REST APIs. As a connection between backend and frontend, we have to take more consideration about the theory of mechanism of REST, and that is what we are learning and what we lack.

Existing API

  • The following graph contains APIs which will apply on the project but not necessary to be restricted by them.


  • In addition, we will also consider some APIs of web-service in the spreadsheets.

Deployment

  • Snip of the deployment on the mobile-phone

  • The interface of the app

  • Containing 5 tab pages



  • Login page


  • Register Page


  • Showing loading


  • Successfully log in


  • The home page for testing



Project Management


### status report
  • work completed:
    • The basic elements in the web with angular and mobile-application with ionic
    • Research of potential APIs
    • Direction of the future
  • Responsibilities
    • Everyone is working good. No evaluation needs to be explicitly made, when everyone was working so hard in the team
    • No explicit split, but we got contributions from every member.
  • Time taken (#hours)
    • Flex time management
    • Online contact 2hr/week
  • Work to be completed:
    • There is still a long way to go, that is just the beginning
    • Responsibility - As the leader, I will manage all my members to move forward and make progress. - Khalid will help Tian more on model deployment - Chen will focus on APIs stuff - Corey will focus on the provider and the global unit-testing
    • Time to be taken is hard to be estimated.




Bibliography

Link

⚠️ **GitHub.com Fallback** ⚠️