Comments (1)
Hi,
We use some cg conventions (center of camera, look at direction) to describe the camera poses for lego data. This is also one of the default with pytorch3D's setup.
When dealing with real world data, we use COLMAP to get poses (same with fern or room data). The poses and bounds are stored in the poses_bounds.npy and is supported by our code release. If you are using fern, room or other COLMAP data, just set dataset_type: 'real' in your config file (config/legow.yaml). And if you go to dataset.py you will see two different methods to load synthetic data and COLMAP data structure. Setting dataset_type in config will switch between them.
One concern with using COLMAP data is that in current realease, I hardcoded the near and far bounding plane to best constrain the scene. When generate poses from COLMAP, you probably want to manully check the scene depth in COLMAP to determine where is the near and far bounding plane, and manually set it up in line 547-548 in raysampler.py.
We are not using those dataset provided by NeRF because we need enough depth activation, i.e. observe the same scene at very distinctive range, to robustly estimate the parameters. Such variation on depth is missing from those fern or room data so we collected/synthesized our own.
from neuralsea.
Related Issues (9)
- issue with pytorch3d version HOT 3
- TypeError: __init__() got an unexpected keyword argument 'min_depth' HOT 9
- About the camera coordinate type HOT 10
- About WaterNeRF HOT 1
- issues with pip install "git+https://github.com/facebookresearch/pytorch3d.git"
- visdom
- run "python3 train_nerf.py --config-name real_watertank" happen that HOT 1
- License? HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from neuralsea.