...
Analysis - Design B
Efficiency
The efficiency has greatly improved from design A. The user does not have spend time taking a picture for each product he/she would like to view. There is no pre-review selection page; objects detected are automatically loaded as products the user is interested in. However, the tracing arguably takes longer to annotate an object than the touch and drag method. There is no obvious way to modify the outline of the object - the user would have to restart tracing all over again. Apart from the more complicated annotation method, the efficiency of the application has improved by making decision making more simple.
The efficiency issue of comparing a rare or usually unimportant detail still remains: the user would have to travel back and forth between detail pages. Other use cases that this design functions poorly in is if the user has to rearrange the products so that he/she can take a picture with all of the products inside the frame. In this case, it would be easier to just take a multiple of pictures. This design does not allow for multiple pictures to be in the same compare screen.
Learnability
Learnability has decreased in some respects compared to design A. The user must innately know to take a picture that incorporates several objects and annotate them in the same frame, or accidentally learn that the application allows for this. The user can also learn about this feature by watching other users. This is the only feature that is not obvious, and small instruction blurbs 'draw to annotate product(s)' would benefit he learnability of the application. We could also add certain affordances, like animations drawing sample figures around objects. Otherwise, the user should not have a hard time learning that certain buttons on the compare screen bring up more details about that product.
Safety
Design C
Rather than take a picture of the bike, she simply takes a picture of it’s barcode. This does not require any annotation.
...