Skip to content

Latest commit

 

History

History
46 lines (33 loc) · 2.47 KB

README.md

File metadata and controls

46 lines (33 loc) · 2.47 KB
Raw

Interactive Hologram

Holograms have been a dream of tech and entertainment companies as well as consumers ever since a long time. The real hologram can be seen without the aid of any special lenses,screens or props. But there are holographic images produced by means of lenses to create stage illusions like Pepper’s Ghost which are not true holograms, they are called holographic illusions. We will be using Pepper’s Ghost Pyramid to create our holographic illusion.

The DIY hologram pyramid is a simple device that can be made by manipulating a sheet of plastic into the shape of a pyramid with its top cut off. The device creates a 3D-like illusion for the viewer and makes an image or video appear as if it were in midair.

Objective

The objective of the project is to make use of the key points of the hand generated by the SRHandNET model to create a new model which will recognise the hand gestures from the key points. Using these gestures we will assign different actions to the which will be used to manipulate the 3D objects in our hologram.

Working

Interactive Hologram Demo: https://www.youtube.com/watch?v=MbDPa4L7WHs

Hologram Hardware can be constructed according to https://www.youtube.com/watch?v=MrgGXQvAuR4

Reqirements

Tested on CUDA-enabled GPU (Nvidia GTX 1050 and 1660Ti). Code makes use of CUDA

  • PyTorch
  • Tensorflow 2.x
  • OpenCV 4.2.0
  • Pandas
  • SocketIO
  • Matplotlib
  • Seaborn
  • Numba
  • Sklearn
  • CUDA Toolkit 10.x

requirements.txt has all packages from the conda environment

Running the code

Project by:

References:

Y. Wang, B. Zhang and C. Peng, "SRHandNet: Real-Time 2D Hand Pose Estimation With Simultaneous Region Localization," in IEEE Transactions on Image Processing, vol. 29, pp. 2977-2986, 2020, doi: 10.1109/TIP.2019.2955280. https://www.yangangwang.com/papers/WANG-SRH-2019-07.html