Sid–the Voice UI smart mirror that helps improve public speaking skills
Scope: 4 weeks
Categories: Interaction Design, Voice UI
Tools: Sketch, Photoshop, Aftereffects
Advisors: Joshua Musick (Director of Interaction Design, Digital Accelerator at United Technologies), John Leonard (Technology Director at Frog)
Team: Crystal Wang, Margarita Yong, Abhinav Sircar, Ke Hu
Role: UX research, competitive analysis, concept ideation, design iteration, prototype testing, pitch deck
Public speaking fear has 10% impairment on wages, 10% on college graduation and 15% on promotion to management [i]. However, only 8% of those who have public speaking fear seek professional help.
How might we help people improve their public speaking skills?
A Voice UI smart mirror that helps improve public speaking skills by providing real time feedback and progress tracking.
*Why Voice UI: No smudging (on the mirror) & adopting speech recognition technology
Professionals that have public speaking fear which hinders their professional growth.
People that are starting careers which now focus on public speaking.
Students who need the soft skills of presenting for their first jobs.
MOST COMMON MISTAKES
Posture & Body Language
REAL TIME FEEDBACK
MODES OF SID
Hibernating/ activated / talking/ listening / processing/ recording/ paused.
Provides feedback over multiple practices, also option to watch the recorded video.
THE EVOLUTION OF SID
Sid prompts the user with a question.
The possible commands are displayed in quotation marks.
Result: This felt more of a screen interface rather than voice UI.
Re-organized the content and structure in 3 sections (from top to bottom):
Main interaction: user speaks & receives feedback
Added a menu where users can turn on/off each live feedback.
Hand gesture incorporated.
Result: The user needed to pay attention to a lot of things happening on the screen, on top of memorizing the speech.
Simplified the layout and content on screen to avoid cognitive overload.
Replaced option to turn on/off filters with Voice Command, and just left them ON on the screen.
Removed hand gesture.
Added visual representation of Sid, moved him to the middle screen.
Result: Sid located in the middle of the screen interfered with other activities. Real time feedback required lots of attention, which still caused cognitive overload.
Color, texture, shape
How it looked in the mirror
Different modes of Sid (e.g. when Sid is listening vs. talking vs. recording vs. paused).
For recording mode, explored options for adding red–universal recognized color.
THE EVOLUTION OF REAL TIME FEEDBACK
VERSION 1 & 2
Test: Sid only gives feedback when the user is doing something wrong, with 3 states: first warning, second warning, confirm when it’s correct.
- Emojis were not universally understood.
- Took time to process text & emojis at the same time.
- The menu was also a factor of distraction.
- Played around with colors & interactions following the Car Dashboard approach.
- Removed the Menu to clear out the space (then realized Sid position became a distracting element).
- Permanent live feedback.
Results: still couldn’t reduce the cognitive overload.
- Moved Sid to the top and removed feedback Menu to reduce cognitive overload and distractions.
Results: All the three elements on the screen were in motion at the same time, which caused distraction to the user. Also needed to consider visually showing different modes of Sid (listening, talking, recording, paused).
THE EVOLUTION OF TRACKING PROGRESS
© Ngoc Tran 2019