Occupancy Network based 3D Image Reconstruction usingSingle-Depth View

(1) Architecture

(2) Sample Results

(3) Data

(4) Requirements

python 2.7.6

tensorflow 1.2.0

numpy 1.13.3

scipy 0.19.0

matplotlib 2.0.2

skimage 0.13.0

(5) Run

Training

source venv/bin/activate
python main_3D-RecGAN++.py

Test Demo (Download model first)

python demo_3D-RecGAN++.py

Web Application Run Locally

source venv/bin/activate
python main.py

(6) Citation

@inProceedings{Yang18,
  title={Dense 3D Object Reconstruction from a Single Depth View},
  author = {Bo Yang
  and Stefano Rosa
  and Andrew Markham
  and Niki Trigoni
  and Hongkai Wen},
  booktitle={TPAMI},
  year={2018}
}

System Requirements:

Google Cloud Platform - Deep Learning VM Input Framework: PyTorch 1.4 + fast.ai 1.0 (CUDA 10.0) GPU: 1, nvidia-tesla-k80

We have referred to the code provided by the following citation for our implementation:

@inproceedings{Occupancy Networks,
    title = {Occupancy Networks: Learning 3D Reconstruction in Function Space},
    author = {Mescheder, Lars and Oechsle, Michael and Niemeyer, Michael and Nowozin, Sebastian and Geiger, Andreas},
    booktitle = {Proceedings IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)},
    year = {2019}
}

Our modifications to the original architecture is documented within the code.

Installing required packages using anaconda3, run the following commands:

conda env create -f environment.yaml conda activate mesh_funcspace

Compile the required extension modules:

python setup.py build_ext --inplace

Download the required pre-processed dataset (will be downloaded to data/ShapeNet folder)

Run in occupancy_networks_code directory.

bash scripts/download_data.sh

Training:

python train.py <CONFIG_FILE_DIRECTORY/onet.yaml>

Generation using pre-trained model:

python generate.py <CONFIG_FILE_DIRECTORY/onet_pretrained.yaml

souravchk25 / csci5922project Goto Github PK

csci5922project's Introduction

Occupancy Network based 3D Image Reconstruction usingSingle-Depth View

(1) Architecture

(2) Sample Results

(3) Data

Part 1: {ShapeNetCore.v2: bench, chair, couch, table}, 20G

Part 2: {ShapeNetCore.v2: airplane, car, monitor, faucet, guitar, gun}, 9.3G

Real Dataset: {Kinect: bench, chair, couch, table}

(4) Requirements

(5) Run

Training

Test Demo (Download model first)

Web Application Run Locally

(6) Citation

Installing required packages using anaconda3, run the following commands:

Compile the required extension modules:

Download the required pre-processed dataset (will be downloaded to data/ShapeNet folder)

Run in occupancy_networks_code directory.

Training:

Generation using pre-trained model:

csci5922project's People

Contributors

Watchers

Forkers

Recommend Projects

Recommend Topics

Recommend Org