mirrors
/
deep-person-reid
mirror of https://github.com/KaiyangZhou/deep-person-reid.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

update readme

pull/249/head
KaiyangZhou 2019-08-26 11:14:54 +01:00
parent ecaa5ac4bd
commit 6776543683
1 changed files with 5 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
README.rst
View File

@ -155,13 +155,12 @@ The config file sets Market1501 as the default dataset. If you wanna use DukeMTM
-t dukemtmcreid \
--transforms random_flip random_erase \
--root $PATH_TO_DATA \
--gpu-devices 0
--gpu-devices 0 \
data.save_dir log/osnet_x1_0_dukemtmcreid_softmax_cosinelr
The code will automatically (download and) load the ImageNet pretrained weights. After the training is done, the model will be saved as "log/osnet_x1_0_market1501_softmax_cosinelr/model.pth.tar-250".
The code will automatically (download and) load the ImageNet pretrained weights. After the training is done, the model will be saved as "log/osnet_x1_0_market1501_softmax_cosinelr/model.pth.tar-250". Under the same folder, you can also find the `tensorboard <https://pytorch.org/docs/stable/tensorboard.html>`_ file. To visualize the learning curves, you can run :code:`tensorboard --logdir=log/osnet_x1_0_market1501_softmax_cosinelr` in the terminal and visit :code:`http://localhost:6006/` in your web browser.
Evaluation will be automatically performed at the end of training.
To run the test again using the trained model, do
Evaluation is automatically performed at the end of training. To run the test again using the trained model, do
.. code-block:: bash
@ -188,7 +187,7 @@ Suppose you wanna train OSNet on DukeMTMC-reID and test its performance on Marke
--root $PATH_TO_DATA \
--gpu-devices 0
Here we only test the cross-domain performance. However, if you also want to test the same-domain performance, you can set :code:`-t dukemtmcreid market1501`, which will evaluate the model on the two datasets separately.
Here we only test the cross-domain performance. However, if you also want to test the performance on the source dataset, i.e. DukeMTMC-reID, you can set :code:`-t dukemtmcreid market1501`, which will evaluate the model on the two datasets separately.
Different from the same-domain setting, here we replace :code:`random_erase` with :code:`color_jitter`. This can improve the generalization performance on the unseen target dataset.