Code Monkey home page Code Monkey logo

auxilary-depth-aided-semantic-scene-completion's Introduction

ADASSC-Net

Download

3D Ground truth, depth images

  • SUNCG (Test), NYU (Train/Test), NYUCAD (Train/Test) Download
  • SUNCG (Train) Download
  • SUNCG-RGBD (Train/test) Github

2D Labels, RGB

  • NYU Download
  • NYUCAD (No RGB, Labels will be provided soon)

Data details

  • Data format

    1. Depth map : 16-bit png with bitshifting (for visualization). Refer to _read_bitshift() in helper_functions.py for more details.
    2. 3D volume :
      1. 3 floating points store the origin of the 3D volume in world coordinates.
      2. 16 floating points store the camera pose in world coordinates.
      3. 3D volume encoded by run-length encoding. Refer to _read_bin() in helper_functions.py for more details.
  • Data availability

Dataset 3D GT Depth RGB Amodal depth 2D Instance Label 2D Semantic Label
NYU
NYUCAD Use NYU color img
SUNCG
SUNCG-RGB

Data organisation

- ADASSC
  - data
    - temp
      - depthbin_eval.zip
      - nyu_depth_v2_labeled.mat
    - depthbin_eval
      - depthbin
        - NYUtest
        - NYUtrain
        - NYUCADtest
        - NYUCADtrain
        - SUNCGtest_49700_49884
        - ~SUNCGtrain~
      - eval
        - NYUtest
        - NYUCADtest
        - SUNCGtest_49700_49884
  - demo
    - sample_data
    - demo.ipynb
  - scripts

Data preparation

Under ./scripts...

  1. Extract RGB for NYU
    • Run python extract_from_mat.py -t color.
    • Specify -t label for instance and semantic labels.
  2. Depth-to-normal
    • Run python depth2normal.py -d NYU. Refer to _gen_normal() in helper_functions.py for more details.
  3. Depth-to-HHA
    • Run python depth2hha.py -d NYU.
  4. Depth-to-TSDF (CPU ver.) Note: Not recommended for many files as this will take much longer time than GPU-compiled codes.
    • Run python depth2tsdf.py -d NYU -f ../data/depthbin_eval/depthbin/NYUtrain/NYU0003_0000.bin.

3D Data visualization

  • Requires Meshlab Download / Blender Download
  • Visualize TSDF
    • Under ./scripts/vis_utils, run python scene_viewer.py -i path/to/file.npz -p edgenet
    • e.g. Visualizing sample data: python scene_viewer.py -i ../../demo/sample_data/NYU0003_0000.npz -p edgenet. By default, the .ply file will be saved in the same directory as the .npz file

TODO

  • ☐ Add SUNCG-RGB download link
  • ☐ Add NYUCAD labels download link
  • ☐ Add SUNCG-RGB into data organisation tree
  • ☐ Remove old codes

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.