-
-
Gold definition and key understanding prompt
Inspiration
The release of GPT4-V and the potential with wearables.
What it does
It analyzes video from a body worn camera and tries to critique your behavior to try and help you reach goals that you provide it.
How we built it
Body worn camera + GPT4-V + DALLE3 + FFMPEG
Challenges we ran into
Image prompts generated by the model were not great
Accomplishments that we're proud of
Accuracy of vision was very good and the generated feedback was very salient.
What we learned
GPT4-V is very capable of video understanding.
What's next for ReflectiveAI
Add support for longer videos by using scene detection to find key parts of a longer video and then have those processed independently.
SUBMITTED REMOTELY but coming in now and will be in person in thirty minutes
Log in or sign up for Devpost to join the conversation.