Unsupervised-Graph-Association-for-Person-Re-identification
Unsupervised-Graph-Association-for-Person-Re-identification copied to clipboard
The comparision on the image person RE-ID dataset
First, thanks for your excellent work. However, I have a question about the comparison.
Actually, in 4.1, For the image RE-ID datasets, we assume all images of a person in one camera belong to a single tracklet.
It may be unfair to compare with GAN-based, adaptation-based and cluster-based methods? The setting is more like to Intra-Camera Supervised Person Re-Identification: A New Benchmark
.
Hope for your response. Thanks