NVIDIA / flownet2-pytorch

Pytorch implementation of FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks
Other
3.09k stars 739 forks source link

run run_a_pair.py error #198

Open semchan opened 4 years ago

semchan commented 4 years ago

Dear,

I just modified the input images as mine for .jpg as the size are 1280*720, but got the below error. how can I fix it? thanks a lot.

/pytorch/torch/csrc/autograd/python_function.cpp:622: UserWarning: Legacy autograd function with non-static forward method is deprecated and will be removed in 1.3. Please use new-style autograd function with static forward method. (Example: https://pytorch.org/docs/stable/autograd.html#torch.autograd.Function) Traceback (most recent call last): File "run_a_pair.py", line 38, in result = net(im).squeeze() File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 541, in call result = self.forward(*input, *kwargs) File "/media/sem/NDT/VisualTracking/flownet2-pytorch/models.py", line 139, in forward flownetc_flow2 = self.flownetc(x)[0] File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 541, in call result = self.forward(input, **kwargs) File "/media/sem/NDT/VisualTracking/flownet2-pytorch/networks/FlowNetC.py", line 106, in forward concat5 = torch.cat((out_conv5,out_deconv5,flow6_up),1) RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 1. Got 16 and 15 in dimension 3 at /pytorch/aten/src/THC/generic/THCTensorMath.cu:71

lhao0301 commented 4 years ago

Try to resize the image so that it can be divided by 64(net's stride). The problem occurs because of the misalignment within the network(conv, deconv).

semchan commented 4 years ago

noted with thanks

Best Regards 陈尧森,Sem Chan Tel: 15208489326

 

------------------ 原始邮件 ------------------ 发件人: "Hao Luo"<notifications@github.com>; 发送时间: 2020年2月17日(星期一) 中午11:17 收件人: "NVIDIA/flownet2-pytorch"<flownet2-pytorch@noreply.github.com>; 抄送: "Sem chan"<380003859@qq.com>;"Author"<author@noreply.github.com>; 主题: Re: [NVIDIA/flownet2-pytorch] run run_a_pair.py error (#198)

Try to resize the image so that it can be divided by 64. The problem occurs because of the misalignment within the network(conv, deconv).

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

birdflyto commented 4 years ago

Thanks for your answer

---Original--- From: "semchan"<notifications@github.com> Date: Mon, Feb 17, 2020 17:14 PM To: "NVIDIA/flownet2-pytorch"<flownet2-pytorch@noreply.github.com>; Cc: "Subscribed"<subscribed@noreply.github.com>; Subject: Re: [NVIDIA/flownet2-pytorch] run run_a_pair.py error (#198)

noted with thanks

Best Regards 陈尧森,Sem Chan Tel: 15208489326

&nbsp;

------------------&nbsp;原始邮件&nbsp;------------------ 发件人:&nbsp;"Hao Luo"<notifications@github.com&gt;; 发送时间:&nbsp;2020年2月17日(星期一) 中午11:17 收件人:&nbsp;"NVIDIA/flownet2-pytorch"<flownet2-pytorch@noreply.github.com&gt;; 抄送:&nbsp;"Sem chan"<380003859@qq.com&gt;;"Author"<author@noreply.github.com&gt;; 主题:&nbsp;Re: [NVIDIA/flownet2-pytorch] run run_a_pair.py error (#198)

Try to resize the image so that it can be divided by 64. The problem occurs because of the misalignment within the network(conv, deconv).

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

fanbooo commented 3 years ago

any solution?