OutOfTheOrdinary: Demand Analysis

Basic Information of this assignment

Which course does this assignment belong to	EE308
Where are the requirements for this assignment?	Requirement
Team Name	OutOfTheOrdinary
The goal of this assignment	Finish the software requirements specifications and requirements analysis PPT

Team contribution table

Student ID	Work Description	Contribution
832102224黄园奎	Frontend Developer	10%
832102229蔡书翰	Back-end engineer, algorithm engineer	10%
832101112蔡正阳	UI designer	10%
832102226黄锦杰	UI designer	10%
832102207孙振翔	Product manager	10%
832101305林陶然	back-end engineer	10%
832102225黄泽晗	software test engineer， Product manager	10%
832102203王毅婷	PM	10%
832002226马庆宏	UI designer	10%
832102216陈卓然	Project Manager	10%
832102111肖凯翔	Product manager	6%

Links of documents

Requirements Specification	document
Analysis PPT	PPT

1. Introduction

1.1Purpose of writing

The purpose of this project is to realize face recognition roll call, which provides convenience for universities, companies and gatherings roll call. For our team, we need to learn how to design and develop a product from start to finish, and lay a foundation for future professional study and work.

1.2Project background

• Software name: Face Meeting Teaching conference management system based on face recognition
• Project task proposer: Teacher Lou
• Project developer: Out of the ordinary team

1.3References

National standard text of “Software Requirements Specification”

2.Overall description

2.1 Objectives

2.1.1 Development intention

We hope to use face recognition technology to make a roll call and sign software for universities and enterprises, which can accurately and quickly identify all the present personnel, count the current number of people, and determine the absence of personnel, and calculate the attendance rate.

2.1.2Application goals and scope

This software is targeted at institutions and groups such as universities and enterprises that have a demand for attendance.

2.1.3 Product Prospects

At present, the interface and functions of the software application are relatively simple, and we hope to add more functions for users in the future, such as customizing calendars and assembling schedules, and using them as memos and schedules; View all upcoming meetings and announcements on your calendar, and post timely notices to your team about the time of the meeting; Set up chat boxes, communication between team members and meeting organizers, etc., to enhance the competitiveness of the software and thus attract more users.

3. Specific requirements

3.1 Class Diagram

3.2 Properties

3.2.1 Availability

Easy to operate and understand, the interface design is simple and easy to use.

3.2.2 Convenience

The software we designed can recognize the face by taking a photo and correspond it to the face one by one, which is fast, simple and accurate

4. Interface prototype

Interface sketch:

5. Function description and acceptance verification standards

5.1Detailed function description

The screens from left to right are Events, Calendar, Home Page, Team, My.

• 1. The activity interface includes notifications and private messages, which can create bulletin boards, send and receive files, and input dialog boxes
• 2. Calendar interface, you can create itineraries, count the number of itineraries and attendance
• 3. Home page interface, that is, face photo interface. After the statistics are completed, the missing persons will be listed in the page according to the itinerary list, and the user’s picture and name will be displayed; The user can confirm the presence by manually checking the box
• 4. Team interface, that is, the team list, you can view the name, valid time and list of the team; You can also create new teams to operate, etc
• 5. My interface, register or login, fill in the account and password; Display my name, ID, profile picture and face, can be edited; Provide suggestions, feedback, etc

5.2 Functional Acceptance Criteria

Face recognition performance requirements:

• False recognition rate Rejection rate: in the indoor environment, the face recognition module rejection rate is less than $1\%$, and in the outdoor non-direct sunlight condition, the rejection rate is less than $5\%$. In any condition, the error rate is less than $0.1\%$.
• Recognition time: under indoor illumination, when the face recognition module is at $1:N(N<200)$, the recognition time is less than $2$ seconds; In $1:1$ mode, the recognition time is less than $1$ second.
• Recognition Angle: The recognition success rate of the face recognition module is greater than $99\%$ under the upper and lower about $15°$ Angle, and the recognition success rate is greater than $95\%$ under the upper and lower about $20°$ Angle.
• Recognition light intensity: The face recognition module requires that the recognition rate is not less than $95\%$ at $5-10000$lux light intensity.
• Recognition distance: Face recognition module requires recognition distance between $0.2-0.8$M. The recognition rate is not less than $95\%$.

challenges

Due to the core function of this software being facial recognition, the accuracy requirements and avoidance of confusion in facial recognition are significant challenges. How to quickly and accurately recognize faces with biometric features is what we are currently considering. At the same time, we need to consider how to solve the high concurrency requirements brought about by a large number of accesses and the large server computing power required to run neural networks.