search

XLearning-SCU / 2022-CVPR-DART

PyTorch implementation for Learning with Twin Noisy Labels for Visible-Infrared Person Re-Identification (CVPR 2022).
30 stars 7 forks source link

Acquisition of the noise id #4

Closed WheatCao closed 1 year ago

WheatCao commented 1 year ago

Hello, I have some doubts about the part of the code that differentiates TN, and whether the use of rgb_noiseIdx for differentiation is causing data leakage. By definition, the model should not be able to directly obtain which noise ids are included. I'm not sure if I'm wrong about this variable. Can u give me an answer? T = (prob_A_V[evaltrainset.rgb_noiseIdx] < args.p_threshold). sum() / len(evaltrainset.rgb_noiseIdx)

XLearning-SCU commented 1 year ago

Hi,

We just use the posted code for observing and logging. The statistics of TP, FN, FP and TN could help readers better understand the robustness achieved.

Notably, the rgb_noiseIdx and rgb_cleanIdx would not be used for training and testing. Therefore, it is impossible to cause data leakage.

WheatCao commented 1 year ago

Thanks for your answer. I'm sorry I misread it.