Hi, thanks for the code! I am trying to reproduce your result using your code, but I only get a mIoU of 40 after 45000 iterations. The result get even worse after 65000 iterations. I am wondering if it is very sensitive to how many iterations to train? Also is there other things to mind using your code?

class Classifier_Module(nn.Module):
    def __init__(self, inplanes, dilation_series, padding_series, num_classes):
        super(Classifier_Module, self).__init__()
        self.conv2d_list = nn.ModuleList()
        for dilation, padding in zip(dilation_series, padding_series):
            self.conv2d_list.append(
                nn.Conv2d(inplanes, num_classes, kernel_size=3, stride=1, padding=padding, dilation=dilation, bias=True))

        for m in self.conv2d_list:
            m.weight.data.normal_(0, 0.01)

    def forward(self, x):
        out = self.conv2d_list[0](x)
        for i in range(len(self.conv2d_list) - 1):
            out += self.conv2d_list[i + 1](x)
            return out

Why is the “return out " statement on the for loop? In my understanding, out is the sum of the pyramid layer , so it should be outside of the for loop.

Hi @wasidennis, I use your provided res101 backbone and train deeplab v2 on GTA as your code, Then I evaluate the model on cityscapes, but I only got 31.1 which is far from your baseline result of 36.6. Did I miss anything? Looking forward to your reply.

Hi,

Been trying to replicate the gta -> cityscapes results on an AWS instance (p2.xlarge). I am using docker with cuda 8.0 and pytorch 0.4.1. Running the following:

python train_gta2cityscapes_multi.py --snapshot-dir ./snapshots/GTA2Cityscapes_single_lsgan \
                                     --lambda-seg 0.0 \
                                     --lambda-adv-target1 0.0 --lambda-adv-target2 0.01 \
                                     --gan LS

It takes about 5.7 seconds per iteration. Given that the model converges at 120k iterations, it's gonna take me more than a week to train it, which sounds insane. Is there something wrong here or are those the expected times?

Hello. I tried to execute your code, but it gives out of memory error. I am working with NVIDIA Titan Xp, which has 12GB memory capacity. It seems that the input sizes of the source and target images are quite big. ((1280,720) and (1024,512))

Can I ask which GPU devices you use for training?

Hello, It is a great work. I saw the readme file where you say the in pytorch-0.4 version the multi-level learning needs the weights for adv learning to be 0.0005 and 0.00005 respectively. I would like to know how you came up with this value ? I am asking because I am working on different dataset and it would be very helpful if I can understand how to specify the reasonable value for adv learning.

@wasidennis @m3phisto @hfslyc , i am trying to run the code (VGG source only). But when i run as you mentioned in previous thread, with --lambda-adv-target1 0 --lambda-adv-target2 i got this error:"train_gta2cityscapes_multi.py", line 311, in main pred1, pred2 = model(images) ValueError: not enough values to unpack (expected 2, got 1)`

Are there any modifications that have to be done for training source only (without) adaptation experiment? If yes, i would be pleased if you could share the code.

I train my data using this method. The loss_adv1 is increased. loss_seg1 and loss_D1 is decreased. In this situation, should I make LAMBDA_ADV_TARGET1 larger ?

Hi, Thank you for sharing your code. I met a problem which is as follows:

**Traceback (most recent call last):

File "train_gta2cityscapes_multi.py", line 412, in main() File "train_gta2cityscapes_multi.py", line 304, in main loss.backward() File "/home/yaxing/anaconda2/envs/pytorch/lib/python2.7/site-packages/torch/autograd/variable.py", line 156, in backward torch.autograd.backward(self, gradient, retain_graph, create_graph, retain_variables) File "/home/yaxing/anaconda2/envs/pytorch/lib/python2.7/site-packages/torch/autograd/init.py", line 98, in backward variables, grad_variables, retain_graph) RuntimeError: there are no graph nodes that require computing gradients**

Note I never change the code. During training, the error happens instead of the beginning. There are three times(iter:25281/25000, 8500/25000, 600/25000 ). The pytorch is 0.2.0_4 with python2.7.

please forgive me to bother you

Hi, thanks for the great work and the codes. In the original implementation, i only found the output space adaptation, since the both outputs from deeplab_multi are coming from the classifier layers and the upsampled for the later alignment. So if i want to implement the real feature space adaptation, i should take the feature maps right before the classifier layers, right? And also a question, the output of from layer5, in my opinion, is not shallow enough, would aligning the features coming from very shallow layers help the domain adaptation?

looking forward to your reply.

Hello！ I have trained your model successfully recently on GTA5 and Cityscapes. And I want to try it on other dataset. I am a newbie in this domain and I met a problem. The new dataset labels are black-and-white images. I found that there maybe some problems in labels. So I want to know that what does "label2train" mean in json file? I really can not understand these numbers([0, 255], [1, 255], [2, 255], [3, 255], [4, 255], [5, 255], [6, 255], [7, 0], [8, 1], [9, 255], [10, 255], [11, 2], [12, 3], [13, 4], [14, 255], [15, 255], [16, 255], [17, 5], [18, 255], [19, 6], [20, 7], [21, 8], [22, 9], [23, 10], [24, 11], [25, 12], [26, 13], [27, 14], [28, 15], [29, 255], [30, 255], [31, 16], [32, 17], [33, 18], [-1, 255]).

And some of them is applied in the gta5_dataset.py in init. I really can not understand these numbers.
Maybe my problem is very stupid. I am really a newbie just want to learn more. Thank you very much!!

Dear author,

I was using this code for my project and I found the discriminator loss is further divided by 2. I am not very clear with GAN and could you please help explain why to divide 2? https://github.com/wasidennis/AdaptSegNet/blob/fca9ff0f09dab45d44bf6d26091377ac66607028/train_gta2cityscapes_multi.py#L362

Thanks a lot!

hello author,

Did you use the the mirror and scale augmentaions in the dataloader. There are arguments in the dataloader function, but have not been used ?

thank you

when I load the pretrained model which is downloaded from

http://vllab.ucmerced.edu/ytsai/CVPR18/DeepLab_resnet_pretrained_init-f81d91e8.pth

then torch.load() will report an error

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/data1/xxx/software/anaconda3/envs/torch18/lib/python3.8/site-packages/torch/serialization.py", line 593, in load
    return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
  File "/data1/xxx/software/anaconda3/envs/torch18/lib/python3.8/site-packages/torch/serialization.py", line 779, in _legacy_load
    deserialized_objects[key]._set_from_file(f, offset, f_should_read_directly)
RuntimeError: unexpected EOF, expected 278891 more bytes. The file might be corrupted.

Is the online model incomplete？

Thank you to share your amazing code!

I try to use deeper segmentation network. But after changing the segmentation backbone network, the adversarial loss diverges only after a few steps.

Can you share a training log file of single-level adaptation training for both VGG-16 & ResNet-101? Thank you!

Hello i want to train model faster.
Could i decrease the NUM_STEPS to 25000? what are the effects of decreasing the NUM_STEPS ? I will be very grateful if you answer my questions sincerely