Homepage 8 August 2016

LaHacks [Bad Mistakes]

Our Project

A mobile application that utilized the phone’s camera to provide audio feedback to the user about their surroundings. The application was targeted towards the visually impaired to enrich their visual experience. We used machine learning provided by Microsoft Cognitive Services and IBM Watson to analyze the photos that were taken every other second by the app.

What I learned

Up until the live demonstration, our project was doing great. We were met with very positive feedback in the pre-judging stage. Our application did not have troubling bugs and those who tested it really had fun. We initially thought we had a shot at winning the first place prize. In comparison to the other top 10 teams, I still believe today that my team developed the best project. I understood that presentation meant everything but our lack of preparation really hurt us. We were unsure of whether to play a recording of a demo or instead show it live in front of the audience. We ultimately thought a live demonstration would be best. Wouldn’t it be awesome for the audience to experience the application live? In our case, NO! We failed to think of how an on-stage scenario would affect our application; mainly the dimmed background and reflecting stage lights. Not realizing that being on-stage would affect the camera’s ability in capturing quality photos was disastrous. Watch the Video. Much of the phone’s screen was black. The audio feedback was incredibly inaccurate and sometimes 100% wrong. If we had the proper preparation and thought ahead, this could have easily been fixed. We should have instead done is play a recorded video of one team member using the application in daylight.

The major takeaway? PLAN AHEAD.

The team that did win the first place prize? They had no live demo.
Presentation is everything.