Mithunjack / Thesis-NeRFs

Denoising 3D TEM tomography via Advanced Neural Radiance Fields(NeRF)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Thesis-NeRFs 🧩

Denoising 3D TEM tomography via Advanced Neural Radiance Fields(NeRF)

Most Important paper for TEM -> NerfMM 📔

Usefull commands 👽

Linux 🤸

  1. Cude Version nvidia-smi // 11.8
  2. Python Version python3 --version // 3.8.15
  3. Clear cmd reset
  4. Conda path setup path
  5. ubuntu architecture uname -m // x86_64
  6. ubunty version and machine id hostnamectl
  7. nvcc version nvcc -V
  8. to find or locate file locate {nvcc}
  9. pytorch -v pip3 show torch // 1.12.1+cu113
  10. linux background proccessing htop
  11. Kill VS code server Remote-SSH: kill VS Code Server on Host Documentation
  12. Installing stuff without sudo Documentation
  13. Extract file tar -xvf cmake-3.x.x.tar.gz
  14. Debian-based Linux Destribution: cat /etc/os-release
  15. Change GPU export CUDA_VISIBLE_DEVICES=1
  16. For building CMAKE cmake . -B build -DCMAKE_CUDA_COMPILER:STRING="/usr/local/cuda-11.8/bin/nvcc"

NeRFstudio Documentation 📄

NerfStudio ✈️

  1. check available model ns-train --help
  2. torch version check in anaconda pip3 show torch
  3. With a specified websocket port ns-train nerfacto --vis viewer --viewer.websocket-port=7008
  4. Resume training from one certain point ns-train nerfacto --data data/nerfstudio/poster --trainer.load-dir {outputs/.../nerfstudio_models}
  5. Pre processing custom data ns-process-data images --data data/custom_data --output-dir outputs/custom_data_preprocessed --no-gpu

Anaconda 🐍

  1. Creating conda env conda create --name nerfstudio -y python=3.8
  2. Activate env conda activate nerfstudio
  3. active env conda info -e
  4. Jupyter notebook password setup
  5. Remove env conda env remove -n ENV_NAME
  6. Remove all env conda remove --name myenv --all
  7. Env list conda env list

Usefull tools for Literature Review 🛠️

  1. It provides an overview of any article and all related research with the assistance of AI -> Paper Digest.
  2. The AI Research Assistant -> Elicit
  3. An app that visualizes all of the related papers for a specific paper -> Litmap
  4. Similar to Litmap, but using a different linked paper visualization website-> Connected paper
  5. For improved writing efficiency -> Writefull
  6. For creating custom vectors or biologically themed graphics -> Bio render
  7. Similarly, litmap and connected paper provide another option -> scite
  8. Reference Manager -> Zotero
  9. Alternative Reference Manager -> Mandele

About

Denoising 3D TEM tomography via Advanced Neural Radiance Fields(NeRF)


Languages

Language:Jupyter Notebook 65.2%Language:TeX 33.9%Language:Python 0.8%