Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

YouTube

Learning 3D Reconstruction in Function Space

Andreas Geiger via YouTube

Overview

Explore cutting-edge techniques in 3D reconstruction through this comprehensive 50-minute virtual talk given at Oxford. Delve into neural implicit models, including occupancy networks, texture fields, occupancy flow, and differentiable volumetric rendering. Gain insights into recent advancements such as conditional surface light fields, PiFU, convolutional occupancy networks, NeRF, and PointRend. Learn about traditional reconstruction pipelines, model architectures, training methods, mesh extraction, object appearance, and generative modeling. Discover how these techniques apply to object motion, reconstruction from image sequences, and interpolation. Download accompanying slides for a deeper understanding of the presented concepts and their practical applications in the field of 3D reconstruction.

Syllabus

Introduction
Research Group
Goals
Traditional Reconstruction Pipeline
Learning
Output Representation
Model Architecture
Training
Mesh Extraction
Representation Power
Object Appearance
Overview
Textures
Results
Combination
generative modeling
object motion
reconstruction from image sequence
interpolation
question
recent results
representation capacity
takehome messages

Taught by

Andreas Geiger

Reviews

Start your review of Learning 3D Reconstruction in Function Space

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.