Person re-identification has become increasingly popular in the community due to its application and research significance. It aims at spotting a person of interest in other cameras.The discriminant information analysis which transforms the original feature vectors by removing the common information, thus deﬁning the discriminant feature. A novel post-ranking framework for person re-identiﬁcation (an unsupervised post-ranking framework) is proposed toimprove the ﬁrst ranking results and outperforms the state-of-the-art approaches. The analysis of the similar appearances of the ﬁrst ranks can be helpful in detecting, hence removing, and such visual ambiguities. Once the initial ranking is available, content and context sets are extracted. Then, these are exploited to remove the visual ambiguities and to obtain the discriminant feature space which is ﬁnally exploited to compute the new ranking. We demonstrate on two pedestrian benchmarks that by learning a more discriminative representation, our method significantly improves the ﬁrst ranks results.
Keywords : Person Re-Identification, Discriminant, Extracted.