Merge branch 'master' of https://github.com/momower1/PointCloudEngine

README.md

@@ -1,11 +1,10 @@
 ## Overview
 - Two seperate point cloud renderers with support for splatting and phong lighting
-- [Ground Truth Renderer](https://github.com/momower1/PointCloudEngine/wiki/Ground-Truth-Renderer) renders a subset of the point cloud and can input it into a neural network
-- [Octree Renderer](https://github.com/momower1/PointCloudEngine/wiki/Octree-Renderer) builds an octree in a preprocessing step and renders the point cloud with LOD control
-- [PlyToPointcloud](https://github.com/momower1/PointCloudEngine/wiki/PlyToPointcloud) tool for converting .ply files with _x,y,z,nx,ny,nz,red,green,blue_ format into the required .pointcloud format
+- Ground Truth Renderer renders a point cloud with splatting, pull push algorithm or neural rendering pipeline and can compare results against a mesh
+- Octree Renderer builds an octree in a preprocessing step and renders the point cloud with LOD control and splatting
+- PlyToPointcloud tool converts .ply files with _x,y,z,nx,ny,nz,red,green,blue_ format into the required .pointcloud format
 
 ## Getting Started
-- Follow install guide for latest Windows 10 64-Bit release from [Releases](https://github.com/momower1/PointCloudEngine/releases)
 - Drag and drop your .ply files onto _PlyToPointcloud.exe_
 - Adjust the _Settings.txt_ file (optional)
 - Run _PointCloudEngine.exe_
@@ -18,16 +17,16 @@
 - Move the camera so close that the individual splats are visible (depending on the scale this might not happen)
 - Adjust the _Sampling Rate_ in such a way that the splats overlap just a bit
 - Enable _Blending_, look at the point cloud from various distances and adjust the blend factor with so that only close surfaces are blended together
+- Pull-Push algorithm can be configured in a similar way
+- Neural Rendering Pipeline requires loading all three exported networks
 
-## Developer Setup
-- Install the following on your Windows machine at the default install locations
-  - [HDF5](https://www.hdfgroup.org/downloads/hdf5)
-  - [Cuda 10.0](https://developer.nvidia.com/cuda-10.0-download-archive?target_os=Windows&target_arch=x86_64&target_version=10)
-  - [Anaconda 3.7](https://repo.anaconda.com/archive/Anaconda3-2019.07-Windows-x86_64.exe) for all users and add it to PATH
+## Developer Setup (Windows)
+- The following libraries are required, make sure that the CUDA toolkit version and Pytorch version are exactly the same
+  - [Cuda 11.7](https://developer.nvidia.com/cuda-11-7-0-download-archive?target_os=Windows&target_arch=x86_64&target_version=11&target_type=exe_local)
+  - [Anaconda3 2022.10](https://www.anaconda.com/products/distribution#Downloads) for all users and add it to PATH
   - Python environment in Visual Studio Installer (no need to install Python 3.7)
-  - Visual Studio 2019 Extension _Microsoft Visual Studio Installer Projects_
-- Run admin command line:
-  - _conda install pytorch torchvision cudatoolkit=10.0 -c pytorch_
+  - [Pytorch 1.13.1](https://pytorch.org/get-started/locally/) _conda install pytorch torchvision torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia_
+- Update include directories, library directories and post build event paths in Visual Studio PointCloudEngine property pages according to the installation paths
 
 ## Example for supported .ply file
 ```
@@ -55,48 +54,13 @@ end_header
 - View the point cloud in different modes
   - Splats: high quality splat rendering of the whole point cloud
   - Points: high quality point rendering of the whole point cloud
-  - Neural Network: renders neural network output and loss channel comparison
+  - Pull Push: fast screen-space inpainting algorithm applied to point rendering
+  - Mesh: render .OBJ textured mesh for comparison
+  - Neural Network: renders neural network output from the sparse point rendering
 - Compare to a sparse subset of the point cloud with a different sampling rate
-- HDF5 dataset generation containing color, normal and depth images of the point cloud
 - Blending the colors of overlapping splats
 - Phong Lighting
 
-## Neural Network View Mode:
-- Neural network should output the splat rendering from a sparse point rendering input
-- The neural network must be loaded from a selected .pt file with _Load Model_
-- A description of the input/output channels of the network must be loaded from a .txt file with _Load Description_
-- Each entry consists of:
-    - String: Name of the channel (render mode)
-    - Int: Dimension of channel
-    - String: Identifying if the channel is input (inp) or output (tar)
-    - String: Transformation keywords e.g. normalization
-    - Int: Offset of this channel from the start channel
-- Example for a simple description file:
-```
-[['PointsSparseColor', 3, 'inp', 'normalize', 0], ['SplatsColor', 3, 'tar', 'normalize', 0]]
-```
-- When a _Loss Function_ is selected
-    - Loss between two channels (_Loss Self_ and _Loss Target_) is computed
-    - Screen area between these channel renderings can be controlled _Loss Area_
-
-## HDF5 Dataset Generation
-- Rendering resolution must currently be a power of 2
-- There are two modes for the dataset generation with parameters in the _HDF5 Dataset_ tab
-  - Waypoint dataset:
-    - Interpolates the camera between the user set waypoints
-    - Use the _Advanced_ tab to add/remove a waypoint for the current camera perspective
-    - _Preview Waypoints_ shows a preview of the interpolation
-    - _Step Size_ controls the interpolation value between two waypoints
-  - Sphere dataset:
-    - Sweeps the camera along a sphere around the point cloud
-    - _Step Size_ influences the amount of viewing angles (0.2 results in ~1000 angles)
-    - Theta and Phi Min/Max values define the subset of the sphere to be sweeped
-    - Move the camera to the desired distance from the center of the point cloud before generation
-- Start the generation process with _Generate Waypoint HDF5 Dataset_ or _Generate Sphere HDF5 Dataset_
-    - Make sure that the density and rendering parameters for the _Sparse Splats_ view mode are set according to [Configure the rendering parameters](https://github.com/momower1/PointCloudEngine#configuring-the-rendering-parameters) 
-    - The generated file will be stored in the HDF5 directory
-    - After generation all the view points can be selected with the _Camera Recording_ slider
-
 # Octree Renderer
 ## Features
 - Loads and renders point cloud datasets and generates an octree for level-of-detail
@@ -133,4 +97,8 @@ end_header
   - char[3] - normalized normal
   - uchar[3] - rgb color
 
+# Training
+- The neural rendering pipeline is trained in Pytorch
+- Datasets for training can be created within the Engine under the Dataset tab
+
 Copyright © Moritz Schöpf. All rights reserved. The use, distribution and change of this software and the source code without written permission of the author is prohibited.