hjimce / depth-map-prediction Goto Github PK
View Code? Open in Web Editor NEWDepth Map Prediction from a Single Image using a Multi-Scale Deep Network
License: GNU General Public License v3.0
Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
License: GNU General Public License v3.0
========================================================================= Depth Map Prediction from a Single Image using a Multi-Scale Deep Network ========================================================================= Authors: David Eigen, Christian Puhrsch and Rob Fergus Email: [email protected], [email protected], [email protected] Requirements ------------- * theano * numpy, scipy * PIL or Pillow Running the Demo ----------------- The demo loads the depth prediction network, compiles a theano function for inference, and infers depth for a single image. To run: > THEANO_FLAGS=device=gpu0 python demo_depth.py This should create a file called "demo_nyud_depth_prediction.png" with the predicted depth for the input "demo_nyud_rgb.jpg". (Substitute the gpu you want to run on for gpu0). Other Information ------------------ This tree contains code for depth prediction network inference. While there is some code relating to training, much of the training code including most data processing is not provided here. We may release this in the future, however. While developing this project, we made a few modifications in theano not currently part of the main codeline. While the above instructions should work for inference on a current unmodified theano build, it may take up more GPU memory than needed due to use of test values for shape information. The git patch file "theano_test_value_size.patch" is also included and might be used to enable this feature on your own tree.
I tried to predict a depth map for a larger image (2562x1708 pixels). Still the output of the depth map (using test.py
) is 74x54.
Is it possible to have bigger outputs?
I saw that the picture is internally downsampled to 320x240. When I commented that part the program was failing. Is the network capable of handling larger inputs?
Hi,
I trying to compiler this work and finded that there is a parameter called params_dir = 'weights/depth'
in the test.py.
But the whole project didn't exist the file called weights.
So I want to know can you provide this file or we can get this weights by ourselves throught this project?
Regards,
Mathilda.
What is your version of theano,cuda?
How can I training on my own data ?
Thank you .
noll
Does this model can be trained by other data set,because I see that there is some code relating to training.
Hi,where is the demo_depth.py
?tx
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.