Research Article
Deep Binary Representation for Efficient Image Retrieval
Algorithm 1
Optimal hash code generation.
| Input: binary code length , number of categories | | Output: code set , satisfies | | (1) codeset.add(0) | | (2) for () do | | (3) flag=0 | | (4) for do | | (5) if Hamming then | | (6) flag=1 | | (7) break | | (8) if flag==0 then | | (9) codeset.add(i) | | (10) return codeset | | Repeat: Perform the algorithm with ,… until the length of code set is larger than . | | Choose codewords from the code set with largest , that will be the target hash code set. |
|