pierlj / ken-burns-effect Goto Github PK
View Code? Open in Web Editor NEW3D Ken Burns effect and dolly zoom from single image.
License: MIT License
3D Ken Burns effect and dolly zoom from single image.
License: MIT License
Even if I modified the path of helper_math.h, I got the following error.
Can you help me with the " unexpected '{' in field name" ERROR ?
What does that means ?
Traceback (most recent call last):
File "kbe.py", line 181, in
ken_burn_pipe((tensorImage+1)/2, zoom_settings, output_path, pretrained_estim=pretrained_estim)
File "/home/liaowq2/code/ken-burns-effect/utils/pipeline.py", line 118, in call
}, self.objectCommon, self.moduleInpaint)
File "/home/liaowq2/code/ken-burns-effect/utils/common.py", line 213, in process_kenburns
objectCommon['dblBaseline'])
File "/home/liaowq2/code/ken-burns-effect/utils/common.py", line 515, in render_pointcloud
'zee': tensorZee
File "/home/liaowq2/code/ken-burns-effect/utils/common.py", line 295, in preprocess_kernel
'''.format(path_to_math_helper) + strKernel
ValueError: unexpected '{' in field name
Hello thanks for sharing this project!
can you assist by sharing the trained weights - se we can test it easily?
hi,@pierlj
i read the readme file and know use pretrained_estim=True to set cv2.cvtColor(numpyImage, cv2.COLOR_BGR2RGB) when use the model from original paper,but i doubt:
1、the original code do not have to use cv2.cvtColor(numpyImage, cv2.COLOR_BGR2RGB);
2、your train code use cv2.cvtColor(numpyImage, cv2.COLOR_BGR2RGB) on data_loader.py,so when inference it should have to use it, but on your kbe.py code it is not;
thanks!
when i run kbe.py, meet the following error
fatal error: /home/s182169/Master_thesis/GAN-Burns-Effect/utils/helper_math.h: no such file or directory
Hi, thanks for your incredible work!
I wanna train the network by myself, but I am confused about how to setup data.
In data_loader.py, it seem that I should download imagenet dataset.
Can you provide the link of which imagenet dateset you use and how to setup the directory structure of imagenet dataset?
Thanks!
Hi, sorry for bothering you again.
As you answered in my previous issue, there are some hardcoded values inside your script.
I have tried using your kbe.py
to demo random pictures on the Internet, and they seem fine without changing any parameters, so it came to me a question.
In utils/pipeline.py
line 26 and 27, you set objcommon's focal and baseline to be 512 and 120 for demo and these match the setting of DIML-outdoor dataset, so I guess you were demoing that dataset.
However, if I want to demo some pictures in other datasets, or the images I took, is it a must to know the focal length and baseline of the camera used to take the images?
Thanks for your helping!
hi,
I want to train estimation。
I found 'path': '/scratch/s182169/GTAV_1080/', in train.py.
do you have a new dataset?
Can you share it with me?
Can you post the dataset, please? thank you
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.