Zhedong Zheng comments

Results 584 comments of


                                            Zhedong Zheng

pre-trained model

在google drive 可能需要梯子

pre-trained model

我这边看链接是 Ok哒～

你好，想问一下，怎么利用训练好的PCB网络模型去提特征，还有如果我想改part的个数，怎么改呢

你好你可以在model.py 里面修改 npart的个数。测试的时候需要把特征concat起来。

Typo

Thanks.

Can't find PCB_test and some other pre-trained models in Google drive

Hi @GayatriPurandharT Recently I update models, so some test models are missing. It would be good if you could re-train the model by yourself. Actually, it usually only takes one...

Making the demo live

Hi @chetansarmak Generally, it will affect accuracy, since the model is trained in a different environment from your testing situation.

模型训练过程中的模型权重更新逻辑

@finger-monkey We find the validation set could not reflect the performance accurately, so we save the last model by default.

"Detected call of `lr_scheduler.step()` before `optimizer.step()`. " in train.py

Thanks @ngalalalekha32 It is due to the new Pytorch warning, and it will not hurt the result.

demo on images other than dataset

Hi , for the custom dataset, you need to extract the feature first by `test.py` and then you could use `demo.py`.

demo on images other than dataset

Yes. @TheLeopardsH I suggest you to check the filename and organization in the Market-1501 dataset.