Systematic confidence #2
Description
We should implement a system so that we can deduce how confident we are about a finger's pose. However, the approach on how to do so is currently unclear.
Should we consider training an AI model to determine if a flex sensor can be trusted or not?
Should we consider using statistics to determine if a flex sensor can be trusted or not?
How do we handle the fact that some flex sensors incorrectly read the splay of a finger as curl?