This is part of my journey of learning NeRF.

2.1. Network Architecture

1. Input Encoding

Similar as NLP, they use position encodings. Like Sinusoid functions. I also remember an encoding method which takes into consider of the 光线的散射

I am currently burying myself into the sea of NeRF. I plan to archive my learning notes here. I am still a beginner, so the notes absolutely contain errors, and are not finished yet.

Contents

Learning NeRF: Reading list, learning references, and plans

Notes of CVPR22' Tutorial:

This is part of my journey of learning NeRF.

2.4. Prior-based reconstruction of neural fields

Sounds like a one-shot task: instead of fitting and optimizing a neural field each for one scene; let's learn a prior distribution of neural field. Then, given a specific scene, it adjusts the neural field in just one forward.

Reference: https://stackoverflow.com/questions/48302810/whats-the-difference-between-hidden-and-output-in-pytorch-lstm

I was so confused when doing a homework on implementing the Luong Attention, because it tells that the decoder is a RNN, which takes \(y_{t-1}\) and \(s_{t-1}\) as input, and outputs \(s_t\), i.e., \(s_t = RNN(y_{t-1}, s_{t-1})\).

But the pytorch implementation of RNN is: \(outputs, hidden\_last = RNN(inputs, hidden\_init)\), which takes in a sequence of elements, computes in serials, and outputs a sequence also.

I was confused about what is the \(s_t\). Is it the \(outputs\), or the \(hidden\_states\)?

论文地址：https://shihmengli.github.io/3D-Photo-Inpainting

作者：Meng-Li Shih, Shih-Yang Su, Johannes Kopf, Jia-Bin Huang

发表： CVPR2020

链接： https://github.com/vt-vl-lab/3d-photo-inpainting

论文地址：https://arxiv.org/abs/2112.13715

作者：Ailing Zeng, Lei Yang, Xuan Ju, Jiefeng Li, Jianyi Wang, Qiang Xu

发表： ECCV 2022

链接： https://github.com/cure-lab/SmoothNet

失眠的话

夜晚就变得绵长

第几节脊椎上了发条，拧动时

奇异的对偶的诗流向新的小节

论文地址：https://arxiv.org/abs/2203.12575

作者：Jiang, Wei and Yi, Kwang Moo and Samei, Golnoosh and Tuzel, Oncel and Ranjan, Anurag

发表： ECCV22

开源代码： https://github.com/apple/ml-neuman

According to https://dev.to/rubberduck/using-usb-with-docker-for-mac-3fdd, it seems impossible to forward USB to a Docker container on MacOS, coz the Docker is running in a virtual environment via hyperkit.

First of all, ports of the host (i.e., MacOS) cannot be directly accessed by any virtual environment (i.e., Docker) on it. So, "you first have to expose it to the virtual machine where Docker is running". However, Docker is running on hyperkit, which doesn't support usb forwarding.

The author provided another way to do it, that is to usedocker-machine, which uses a Virtualbox VM to host the dockerd daemon, to replace the original docker. Then... why bother still using Docker, instead of just using Virtualbox to run the seperated environment?

论文地址：https://tcwang0509.github.io/vid2vid/paper_vid2vid.pdf

作者：Ting-Chun Wang, Ming-Yu Liu, Jun-Yan Zhu, Guilin Liu, Andrew Tao, Jan Kautz, Bryan Catanzaro

发表： NeurIPS 2018

Project：https://tcwang0509.github.io/vid2vid/