eti-p-doray / unet-gan-matting Goto Github PK

View Code? Open in Web Editor NEW

293.0 293.0 82.0 141.09 MB

Background Removal based on U-Net

License: MIT License

Python 100.00%

unet-gan-matting's People

Contributors

Stargazers

Watchers

Forkers

inkimage tuyaliang tannerversety neuwangmeng am1n3e godweed nnnnew jackemuk stanxii junaid199f peardox devarshi87 unitingcoders coorhair rowellpica heniitang vmdao kp-forks imrankk jjp9624022 hlyu368 xlnrp griffin98 mkharibalaji traveldriffter wanglc2008 chrislee2012 jingmouren blaxe05 kuan-li yangtong1989 smokeelow wei-ancoris iashishverma keshav47 rudraina kqw94 nguyenvuducthuy noderaider 87ssfantasy hmate9 msiliu buiduchanh artifacts bcc0910 kylebarz jaxn quyenbui vonzunlei chenhuayou killsking danielwurgaft jiewei119 ottawaer ophoperhpo amitshah makeyoubox marcojoao vaibhavbaswal95 iampaopaoyu kedzior-io syed-cbot bryan-bai focox ayushsojitra sep-eg csie-lab-aug2021 sts0mrg0 mrsandmanrus jinwook-shim deanhnter crazbot kenakai16 phoopyae yepeesteve ileaderx pingpong-ai-models nagyist zncu97 background-removal-expert

unet-gan-matting's Issues

Download a U-net implementation and adapt it to our needs

https://github.com/jakeret/tf_unet

Is there anyone can download the dataset?

I follow the instructions completely to download the dataset, but it failed. The error as follows:

selenium.common.exceptions.WebDriverException: Message: unknown error: net::ERR_CONNECTION_TIMED_OUT
(Session info: headless chrome=89.0.4389.82)

I understand this was probably a project for your personal use, but I found this approach for background removal much different from other better known ones like Segmentation, etc. and wanted to give it a try. Is eval/running it only possible if we have a Trimap and can I not just use any of my normal images to remove its background (I dont have trimaps for them).

I didn't find the GAN part in the project

In the beginning, I used the model I trained to eval found that the picture output was too pink. I thought maybe the dataset I trained was too little, but I used the 83256 you provided to eval the result is also pink, just like the figure 6(a) in your paper when GAN is not used.

Not able to download images

Looks like google image search changed some meta information , so google image download is broken now.
Anywhere I can find the dataset ?
Thanks in advance

Time required for training

How much time does it require to train this model?

[Paper] Intro/related work

Error message running Training

I'm getting an error message when running the Training.

C:\unet-gan-matting-master>python train.py scripts/data
Traceback (most recent call last):
File "train.py", line 13, in
from script import resize
ModuleNotFoundError: No module named 'script'

Am I missing something?

Item no.: 1 --> Item name = portrait transparent background Evaluating... Looks like we cannot locate the path the 'chromedriver' (use the '--chromedriver' argument to specify the path to the executable.) or google chrome browser is not installed on your machine (exception: Message: session not created: This version of ChromeDriver only supports Chrome version 81 )

Item no.: 1 --> Item name = portrait transparent background
Evaluating...
Looks like we cannot locate the path the 'chromedriver' (use the '--chromedriver' argument to specify the path to the executable.) or google chrome browser is not installed on your machine (exception: Message: session not created: This version of ChromeDriver only supports Chrome version 81
)

How to resolved ? .. thanks

train.py: module 'tensorflow' has no attribute 'get_variable'

I got this error when running train.py

2019-10-14 23:49:24.635561: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_100.dll
Traceback (most recent call last):
  File "train.py", line 226, in <module>
    main(args)
  File "train.py", line 72, in main
    train_ids = tf.get_variable('train_ids', initializer=ids[0:split_point], trainable=False)
AttributeError: module 'tensorflow' has no attribute 'get_variable'

I'm running windows 10 and I tried python 3.5 and 3.6 with requirements-gpu.txt installed
I think you specified wrong tensorflow version.

I tried replacing strings like that

    # train_ids = tf.get_variable('train_ids', initializer=ids[0:split_point], trainable=False)
    # valid_ids = tf.get_variable('valid_ids', initializer=ids[split_point:len(ids)], trainable=False)
    # global_step = tf.get_variable('global_step', initializer=0, trainable=False)
    train_ids = tf.Variable(tf.ones(ids[0:split_point]), name="train_ids")
    valid_ids = tf.Variable(tf.ones(ids[split_point:len(ids)]), name="valid_ids")
    global_step = tf.Variable(tf.ones(0), name="global_step")

But I got another error

2019-10-14 23:54:15.875251: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_100.dll
2019-10-14 23:54:17.429887: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library nvcuda.dll
2019-10-14 23:54:17.553446: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1618] Found device 0 with properties:
name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate(GHz): 1.8475
pciBusID: 0000:1f:00.0
2019-10-14 23:54:17.560973: I tensorflow/stream_executor/platform/default/dlopen_checker_stub.cc:25] GPU libraries are statically linked, skip dlopen check.
2019-10-14 23:54:17.566947: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1746] Adding visible gpu devices: 0
2019-10-14 23:54:17.572106: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2019-10-14 23:54:17.582708: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1618] Found device 0 with properties:
name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate(GHz): 1.8475
pciBusID: 0000:1f:00.0
2019-10-14 23:54:17.588355: I tensorflow/stream_executor/platform/default/dlopen_checker_stub.cc:25] GPU libraries are statically linked, skip dlopen check.
2019-10-14 23:54:17.593767: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1746] Adding visible gpu devices: 0
2019-10-14 23:54:18.321282: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1159] Device interconnect StreamExecutor with strength 1 edge matrix:
2019-10-14 23:54:18.326692: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1165]      0
2019-10-14 23:54:18.330506: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1178] 0:   N
2019-10-14 23:54:18.334713: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1304] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4708 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:1f:00.0, compute capability: 6.1)
Traceback (most recent call last):
  File "train.py", line 228, in <module>
    main(args)
  File "train.py", line 75, in main
    train_ids = tf.Variable(tf.ones(ids[0:split_point]), name="train_ids")
  File "C:\Users\ivan\Desktop\unet-gan-matting-master\vendor\python\lib\site-packages\tensorflow_core\python\ops\array_ops.py", line 2583, in ones
    output = fill(shape, constant(one, dtype=dtype), name=name)
  File "C:\Users\ivan\Desktop\unet-gan-matting-master\vendor\python\lib\site-packages\tensorflow_core\python\ops\array_ops.py", line 171, in fill
    result = gen_array_ops.fill(dims, value, name=name)
  File "C:\Users\ivan\Desktop\unet-gan-matting-master\vendor\python\lib\site-packages\tensorflow_core\python\ops\gen_array_ops.py", line 3599, in fill
    _six.raise_from(_core._status_to_exception(e.code, message), None)
  File "<string>", line 2, in raise_from
tensorflow.python.framework.errors_impl.InvalidArgumentError: dims must be a vector, got shape [6069,2] [Op:Fill] name: ones/

How long it did you take to train the modal?

Facing issues while using it in colab

After some experimentation, I found that chromedriver doesn't work in colab. I used !./chromedriver in colab cell

The result was
src/tcmalloc.cc:283] Attempt to free invalid pointer 0x1951c81affc0

But when i try locally ./chromedriver :

`
Starting ChromeDriver 78.0.3904.105 (60e2d8774a8151efa6a00b1f358371b1e
0e07ee2-refs/branch-heads/3904@{#877}) on port 9515
Only local connections are allowed.
Please protect ports used by ChromeDriver and related test frameworks
to prevent access by malicious code.

Any suggestions what can be done ??
I have tried
!sudo apt-get install libjemalloc-dev !apt install libtcmalloc-minimal4
`

eti-p-doray / unet-gan-matting Goto Github PK

unet-gan-matting's People

Contributors

Stargazers

Watchers

Forkers

unet-gan-matting's Issues

selenium.common.exceptions.WebDriverException: Message: unknown error: net::ERR_CONNECTION_TIMED_OUT (Session info: headless chrome=89.0.4389.82)

Recommend Projects

Recommend Topics

Recommend Org

Jobs

selenium.common.exceptions.WebDriverException: Message: unknown error: net::ERR_CONNECTION_TIMED_OUT
(Session info: headless chrome=89.0.4389.82)