Lâm Tuấn Anh
Lâm Tuấn Anh
Actually I found out that it just doesn't run when I switch default `useLexicon` in `demo.lua` to `true` no matter the lexicon file
Those files should be generated when you run test.py
At the time, the author of the paper had not released the code so I decided to code this myself following the implementation details shared in the paper, so it...
I don't remember the details, but when I wrote that part, I wasn't satisfied with the matmul operation (I don't remember the reason). So I had to an element-wise multiplication...
You should be able to find the link to download the files in README. If you want to generate the files yourself just initiate the model with pretrained weights then...
Softened labels are simply indices, each instance would have a different label. The Market1501Dataset class automatically returns the softened labels.
@bainianjie Hi, thank you for your interest in this repo. I wrote an answer to a similar question which you can reference here: https://github.com/ryanaleksander/softened-similarity-learning/issues/1 Feel free to ask if you...
Hi, could you show me traceback? Also, please don't hold back asking any questions you have, I'll try to answer them all.
It seems that python cannot find the config group 'Train Data' in your config file. It looks like this. ``` [Train Data] path=Market-1501-v15.09.15 batch_size=8 memory_vectors_path=memory_table/vectors.pth memory_parts_path=memory_table/parts.pth memory_cameras_path=memory_table/cameras.pth lambda_c=0.02 lambda_p=0.5 reliable_num=4...
Apparently, configparser seemed to have failed to read the config file, most likely because the path was wrong (configparser won't tell you whether or not the file is valid or...