tsaishien-chen / span Goto Github PK

Hello, thank you for open-sourcing the code for your paper. I don't seem to locate the code for Part Feature Extraction and CPDM. And could you point me to where the re-identification is happening?

Thank you

Hi, I saw that your project "is only supported on VeRi-776 dataset currently." But in your paper, you've test mAP on CityFlow dataset. What should I do in order to reproduce on CityFlow dataset?
Thanks for helping!

大佬，可以把切割的模型放出来吗

Congratulations on your success. You did an awesome job in Re-ID, and I got much inspiration from your work.
When would you like to release the code to the public? I would cite your work with my highest respect.

Hello,
When trying to visualize the output, GrabCut and DL rows are turning out to be blank. Can anyone help me out on this?

Thanks

Excuse me.I am a beginner.
I want to know where the triplet loss is in the code.

Paper:
In the second stage, we optimize the rest of our network (Fig. 2 (b)(c))
with two common re-ID losses while SPAN is fixed. The first one for metric
learning is the triplet loss (Ltrip) [28], which is calculated based on the weighted
distance introduced in Sec. 3.3. The other loss for the discriminative learning is
the identity classification loss (LID) [41]. The overall loss is computed as follows:
Lstep2 = λtripLtrip + λIDLID. (7)

It seems that u only release the code of SPAN part. I wanna know if the three 'second stage' in Part Feature Extraction share the same parameters?