test Lowe

_{scroll ↓ to Resources}

Note

• simplest way to resolve a problem in image retrieval task, when there will always be the most similar object in the database to the one that was inputted during inference. If so, how to decide whether this is indeed a match or not.
• the idea is that the same object is much more alike to itself then to the next most similar object (but not itself)
  
  where $x$ is input vector, $x_1$ and $x_2$ are potential matches, closest objects from the database. This ratio is ~ 0 if $x_1$ is a correct match and it is ~0 if $x_1$ and $x_2$ are both false positives.
• reliability of this approach is highly dependent on the number of database entries. The more objects are present, the better this ratio works.

Resources

---

Links to this File

table file.inlinks, file.outlinks from [[]] and !outgoing([[]])  AND -"Changelog"