Faster Neural Radiance Fields Inference
The Neural Radiance Fields (NeRF) proposed an interesting way to represent a 3D scene using an implicit network for high fidelity volumetric rendering. Compa...
Posts by Tags
3D
3D Convolutional Neural Network
3D Features
3D Perception
3D Reconstruction
3D semantic segmentation
3D vision
4D convolutional neural network
Blog
Switching to Visual Studio Code
I’ve been using VIM for most of my Ph.D. years and one of the reasons why I stick with VIM is that I could just ssh to a remote server and recover environmen...
Misconceptions about Memory and Good Documentation
Documentation probably is one of the most important tasks that no one has time for. I also overlook the importance as I get swept by a series of projects and...
Classification
Regression vs. Classification: Distance and Divergence
In Machine Learning, supervised problems can be categorized into regression or classification problems. The categorization is quite intuitive as the name ind...
Data Processing Inequality
Data Processing Inequality and Unsurprising Implications
We have heard enough about the great success of neural networks and how they are used in real problems. Today, I want to talk about how it was so successful ...
Dirichlet Process
Documentation
Switching to Visual Studio Code
I’ve been using VIM for most of my Ph.D. years and one of the reasons why I stick with VIM is that I could just ssh to a remote server and recover environmen...
Misconceptions about Memory and Good Documentation
Documentation probably is one of the most important tasks that no one has time for. I also overlook the importance as I get swept by a series of projects and...
Expectation Maximization
Expectation Maximization and Variational Inference (Part 2)
In the previous post, we covered variational inference and how to derive update equations. In this post, we will go over a simple Gaussian Mixture Model with...
Expectation Maximization and Variational Inference (Part 1)
Statistical inference involves finding the right model and parameters that represent the distribution of observations well. Let $\mathbf{x}$ be the observati...
Extension
Pytorch Extension with a Makefile
Pytorch is a great neural network library that has both flexibility and power. Personally, I think it is the best neural network library for prototyping (adv...
Free Form Deformation
Gaussian Process
Learning Gaussian Process Covariances
A Gaussian process is a non-parametric model which can represent a complex function using a growing set of data. Unlike a neural network, which can also lear...
Gaussian Process Regression
Gentle Introduction to Gaussian Process Regression
Parametric Regression uses a predefined function form to fit the data best (i.e, we make an assumption about the distribution of data by implicitly modeling ...
Generative Adversarial Network
Geometric Pattern Recognition
Global Registration
Google Protocol Buffer
Reading protobuf DB in Python
Reading protobuf DB in python
High-dimensional Convolutional Networks
IDE
Switching to Visual Studio Code
I’ve been using VIM for most of my Ph.D. years and one of the reasons why I stick with VIM is that I could just ssh to a remote server and recover environmen...
Information Theory
Data Processing Inequality and Unsurprising Implications
We have heard enough about the great success of neural networks and how they are used in real problems. Today, I want to talk about how it was so successful ...
Instance Segmentation
Discobox: Weakly supervised Instance Segmentation and Semantic Correspondence from Box Supervision
Abstract
Inverse Reinforcement Learning
Loss
Regression vs. Classification: Distance and Divergence
In Machine Learning, supervised problems can be categorized into regression or classification problems. The categorization is quite intuitive as the name ind...
Matrix Calculus
Short Note on Matrix Differentials and Backpropagation
Mathematical notation is the convention that we all use to denote a concept in a concise mathematical formulation, yet sometimes there is more than one way t...
Memory
Misconceptions about Memory and Good Documentation
Documentation probably is one of the most important tasks that no one has time for. I also overlook the importance as I get swept by a series of projects and...
Neural Networks
Data Processing Inequality and Unsurprising Implications
We have heard enough about the great success of neural networks and how they are used in real problems. Today, I want to talk about how it was so successful ...
Neural Radiance Fields
Neuroscience
Misconceptions about Memory and Good Documentation
Documentation probably is one of the most important tasks that no one has time for. I also overlook the importance as I get swept by a series of projects and...
Nonparametric Bayesian
Object manipulation
Pytorch
Pytorch Extension with a Makefile
Pytorch is a great neural network library that has both flexibility and power. Personally, I think it is the best neural network library for prototyping (adv...
Regression
Regression vs. Classification: Distance and Divergence
In Machine Learning, supervised problems can be categorized into regression or classification problems. The categorization is quite intuitive as the name ind...
Representation Learning
Semantic Correspondence
Discobox: Weakly supervised Instance Segmentation and Semantic Correspondence from Box Supervision
Abstract
Variational Inference
Expectation Maximization and Variational Inference (Part 2)
In the previous post, we covered variational inference and how to derive update equations. In this post, we will go over a simple Gaussian Mixture Model with...
Expectation Maximization and Variational Inference (Part 1)
Statistical inference involves finding the right model and parameters that represent the distribution of observations well. Let $\mathbf{x}$ be the observati...
Wasserstein distance
Regression vs. Classification: Distance and Divergence
In Machine Learning, supervised problems can be categorized into regression or classification problems. The categorization is quite intuitive as the name ind...
Weakly supervised learning
Discobox: Weakly supervised Instance Segmentation and Semantic Correspondence from Box Supervision
Abstract
back propagation
Computing Neural Network Gradients
Gradient propagation is the crucial method for training a neural network
barycentric coordinate
Barycentric Coordinate for Surface Sampling
Fast way to sample points on a triangular mesh
caffe
Making a Caffe Layer
Caffe is one of the most popular open-source neural network frameworks. It is modular, clean, and fast. Extending it is tricky but not as difficult as extend...
correspondence
Universal Correspondence Network
TL;DR: Universal Correspondence Network proposed the first fully convolutional way to learn contrastive embeddings for correspondences.
f-divergence
Regression vs. Classification: Distance and Divergence
In Machine Learning, supervised problems can be categorized into regression or classification problems. The categorization is quite intuitive as the name ind...
fully convolutional neural network
Universal Correspondence Network
TL;DR: Universal Correspondence Network proposed the first fully convolutional way to learn contrastive embeddings for correspondences.
graphics
Faster Neural Radiance Fields Inference
The Neural Radiance Fields (NeRF) proposed an interesting way to represent a 3D scene using an implicit network for high fidelity volumetric rendering. Compa...
matrix norm
mesh
Barycentric Coordinate for Surface Sampling
Fast way to sample points on a triangular mesh
multi view
neural network
Ghost of 3D Perception: Permutation Invariance Matters? Convolutions are Permutation Invariant!
Are you familiar with the python dictionary class? Let me give you a quick test to check your level of knowledge.
Faster Neural Radiance Fields Inference
The Neural Radiance Fields (NeRF) proposed an interesting way to represent a 3D scene using an implicit network for high fidelity volumetric rendering. Compa...
Universal Correspondence Network
TL;DR: Universal Correspondence Network proposed the first fully convolutional way to learn contrastive embeddings for correspondences.
Computing Neural Network Gradients
Gradient propagation is the crucial method for training a neural network
point cloud
Barycentric Coordinate for Surface Sampling
Fast way to sample points on a triangular mesh
prediction
programming
Setting Class Attributes in Python
Setting class attributes in python can be tedious. In this post, I want to summarize a trick that I’ve been using to simplify this process.
python
Setting Class Attributes in Python
Setting class attributes in python can be tedious. In this post, I want to summarize a trick that I’ve been using to simplify this process.
Reading protobuf DB in Python
Reading protobuf DB in python
reconstruction
recurrent neural network
rendering
Faster Neural Radiance Fields Inference
The Neural Radiance Fields (NeRF) proposed an interesting way to represent a 3D scene using an implicit network for high fidelity volumetric rendering. Compa...
scene graph
single view
spatio-temporal perception
universal correspondence network
Universal Correspondence Network
TL;DR: Universal Correspondence Network proposed the first fully convolutional way to learn contrastive embeddings for correspondences.