ObstructionRemoval

ObstructionRemoval copied to clipboard

Hi, thanks for sharing this excellent work. I met a checkpoints mismatch problem.  When I want to run the 'test_fence.py' file, I load the checkpoints under the folder shown...

csqiangwen

您好 我是研究這個論文的大學學生 我想請問一開始我和我的組員使用 anaconda prompt 去執行test_reflection.py ， 裝了很多執行軟體後(numpy , pandas ......) 最後還是無法執行 。 想請問你們當初也是用虛擬環境去執行嗎? 還是是用別的方法去跑這些檔案呢? 謝謝! i wonder if anyone could help me with executing the code . Or narrate...

ritasoup

Hi ! Could anyone please help/guide/walkthrough the process of how to ship this obstruction removal project onto AWS SageMaker ? I'm trying to deploy this work as an endpoint so...

Raja-Salman-Tariq

Hi ! I'm reusing the model a number of times, but everytime I do so I use run_reflections.py. I beleive this script causes the entire model to be recompiled. I...

Raja-Salman-Tariq

How much VRAM is needed?

1

I created a collab to use this project with the Googles GPU (I think it comes with a Tesla T4 ~ 16GB VRAM) and I still get OORM errors, any...

filipppp

Google Colab Download Link dead

3

When trying to use the colab example, it tries to get the whole zip from this URL, which is inaccessible https://www.cmlab.csie.ntu.edu.tw/~yulunliu/ObstructionRemoval_/code_and_ckpt.zip

filipppp

Testing on your own data

2

Hi everyone!! I'd really appreciate if anyone had any ideas for how to test this on your own data - is it only possible to test on 5 images at...

Sicily-F

TypeError: nn() got an unexpected keyword argument 'reuse'

6

Hello I have a trouble to test this please help me > CUDA_VISIBLEDEVICES=0 python3 test_fence.py > 544 > 960 > 2020-05-08 15:32:38.519022: E tensorflow/stream_executor/cuda/cuda_driver.cc:397] failed call to cuInit: CUDA_ERROR_NO_DEVICE >...

max96min

Firstly, thank you for sharing the wonderful work. I was about to test how your network works on some scenes with adherent raindrops, but in the ckpt.zip you provided doesn't...

ejhung

According to your method, I took several groups of real pictures by myself and put them into your model. The training separation failed. I would like to ask you whether...

Lottie641