search

emrahbasaran / SPReID

Code for our CVPR 2018 paper - Human Semantic Parsing for Person Re-identification
MIT License
141 stars 32 forks source link

freature campare #3

Open kunfei1 opened 6 years ago

kunfei1 commented 6 years ago

I have some questions about the SPReid. First, I didn't find distance measurement in the article, did you use Euclidean distance to compare the two freature？ Second ,I use the github's code and generate the 6144-dim freature,but the article seem use the 2048-dim freature, I have confused whether I have a mistake understanding and want to confirm about it. Thank you for your time and help, I will really appreciate it!

emrahbasaran commented 6 years ago

Hi, We use Euclidean distance to compute the distance between the feature vectors. In SPReID, we have three 2048-dim feature vectors. The first one is global representation obtained using global average pooling after InceptionV3. The second one is obtained using foreground region, and the last one is constructed using element-wise max operation on the features of the body parts. You can find more detail in Section 3.3 in the paper.

wuxifang1994 commented 5 years ago

I have some questions about the SPReid. First, I didn't find distance measurement in the article, did you use Euclidean distance to compare the two freature？ Second ,I use the github's code and generate the 6144-dim freature,but the article seem use the 2048-dim freature, I have confused whether I have a mistake understanding and want to confirm about it. Thank you for your time and help, I will really appreciate it!

hi, I also get the 6144-dim feature, so how did you solve this problem? I don't know which step is wrong. Thank you for your help.

ajwl-pmli commented 4 years ago

Hello, Do you know how to compute the rank1 and mAP? could you give me the code which compute the rank1 and mAP, My mail is pku1401210454@163.com, Thank you very much! @kunfei1 @wuxifang1994