jiwei0921/conet

Code for ECCV 2020 paper. "Accurate RGB-D Salient Object Detection via Collaborative Learning".

conet's Issues

There is a difference between the target size and the input size in smooth_l1_loss

E:\Salient Object Detection\2020ECCV\CoNet-master\CoNet\trainer.py:79: UserWarning: Using a target size (torch.Size([2, 256, 256])) that is different to the input size (torch.Size([2, 1, 256, 256])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size.
loss4 = F.smooth_l1_loss(high_depth, depth, reduction='sum')
E:\Salient Object Detection\2020ECCV\CoNet-master\CoNet\trainer.py:81: UserWarning: Using a target size (torch.Size([2, 256, 256])) that is different to the input size (torch.Size([2, 1, 256, 256])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size.
loss6 = F.smooth_l1_loss(pred_depth, depth, reduction='sum')

How to solve the problem during training. Thank you very much for your reply

代码支持非英伟达显卡的电脑运行吗？

您好，我是一名对显著性检测感兴趣的小菜鸟，作为自学python小白，希望您可以帮助我，因为安装pytorch时安装的是only cpu 版本（电脑无英伟达显卡，是inter uhd 620），就担心是不是对跑代码有影响，因为pytorch里的package反复报错，

Traceback (most recent call last):
File "demo.py", line 136, in
training.train()
File "C:\Users\conet\trainer.py", line 169, in train
self.train_epoch()
File "C:\Users\conet\trainer.py", line 60, in train_epoch
for batch_idx, (data, target, depth, edge) in enumerate(self.train_loader):
File "C:\Users\anaconda3\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 517, in next
data = self._next_data()
File "C:\Users\anaconda3\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 1182, in _next_data
idx, data = self._get_data()
File "C:\Users\12874\anaconda3\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 1148, in _get_data
success, data = self._try_get_data()
File "C:\Users\anaconda3\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 986, in _try_get_data
data = self._data_queue.get(timeout=timeout)
File "C:\Users\anaconda3\envs\pytorch\lib\multiprocessing\queues.py", line 104, in get
if not self._poll(timeout):
File "C:\Users\12874\anaconda3\envs\pytorch\lib\multiprocessing\connection.py", line 257, in poll
return self._poll(timeout)
File "C:\Users\anaconda3\envs\pytorch\lib\multiprocessing\connection.py", line 330, in _poll
return bool(wait([self], timeout))
File "C:\Users\anaconda3\envs\pytorch\lib\multiprocessing\connection.py", line 859, in wait
ready_handles = _exhaustive_wait(waithandle_to_obj.keys(), timeout)
File "C:\Users\anaconda3\envs\pytorch\lib\multiprocessing\connection.py", line 791, in _exhaustive_wait
res = _winapi.WaitForMultipleObjects(L, False, timeout)

loss7 = cross_entropy2d(pred_sal2, target, weight=None, size_average=self.size_average)

loss7 = cross_entropy2d(pred_sal2, target, weight=None, size_average=self.size_average)
loss_pre = ((loss3+loss5+loss7)+(loss4 + loss6) * 2.5) / 5
该loss会导致错误：
RuntimeError: cuDNN error: CUDNN_STATUS_NOT_SUPPORTED. This error may appear if you passed in a non-contiguous input.

我想知道这里该如何修改？

训练集缺少train_edge

你好，
感谢你的分享。
我下载了你的数据集以后，对网络进行了训练，但是在训练过程中，读取了train_images、train_masks和train_depth，却报错说我的训练集缺少train_edge，请问这个edge文件夹是什么，该如何生成？
谢谢