Sequential and Correlated Image Hash Code Generation with Deep Reinforcement Learning

Image hashing is an algorithm used to represent image with a unique value. Hashing methods, which are generally developed search for similar examples of image, have gained new dimension the use deep network structures and better results started be obtained methods. The models consider hash functions independently do not take into account correlation between them. In addition, most existing data-dependent methods pairwise/triplet similarity metrics that capture data relationships from local perspective. this study, Central metric, can achieve results, adapted reinforcement learning method sequential strategy, successful in binary codes. By taking errors previous model presented performs interrelated central based learning.