ProFORMA: Probabilistic Feature-based On-line Rapid Model Acquisition
QiPanComputerVisionCited 3× on HN, 2010-2014
571K views · 17 years ago · demo · dated
A three-minute demo of ProFORMA, a system from Cambridge that builds textured 3D models live from a video of an object.
HN cites it as an early demo of generating 3D models just by filming an object, in threads on Kinect, 3D printing and depth cameras.
5:513-Sweep: Extracting Editable Objects from a Single Photo, SIGGRAPH ASIA 2013
Daniel CohenOr
2.3M views • 13 years ago
4:17AI Learns Noise Filtering For Photorealistic Videos | Two Minute Papers #215
Two Minute Papers
65K views • 8 years ago
6:36Face2Face: Real-time Face Capture and Reenactment of RGB Videos (CVPR 2016 Oral)
Matthias Niessner
5M views • 10 years ago
20:39AI Language Models & Transformers - Computerphile
Computerphile
354K views • 7 years ago
5:10Dual Photography
mm648230
24K views • 17 years ago
5:25Make Your Renders Unnecessarily Complicated
sirrandalot
1.2M views • 3 years ago
17:403D Gaussian Splatting! - Computerphile
Computerphile
273K views • 2 years ago
1:10Introducing Word Lens
QuestVisual
5.5M views • 15 years ago
8:09Google Earth’s Incredible 3D Imagery, Explained
Nat and Friends
3.1M views • 9 years ago
57:57Lecture 1 | Introduction to Convolutional Neural Networks for Visual Recognition
Stanford University School of Engineering
2.9M views • 9 years ago
4:28Eulerian Video Magnification
Miki Rubinstein
788K views • 14 years ago
6:18A Style-Based Generator Architecture for Generative Adversarial Networks
Tero Karras FI
1.3M views • 7 years ago
5:26Neural Network Learns The Physics of Fluids and Smoke | Two Minute Papers #118
Two Minute Papers
209K views • 9 years ago
16:43Blender 3.0 Beginner Tutorial - Part 1
Blender Guru
9.3M views • 4 years ago
2:07AI Live Painting in Krita - Interactive updates
Acly
46K views • 2 years ago
19:09A Better Default Colormap for Matplotlib | SciPy 2015 | Nathaniel Smith and Stéfan van der Walt
Enthought
177K views • 11 years ago
8:34Enhancing Photorealism Enhancement
ISL and Collaborators
6.3M views • 5 years ago
24:04FreeCAD 0.20 For Beginners | 1 | Setup , preferences and navigation | Easy Tutorial for Learning CAD
MangoJelly Solutions for FreeCAD
636K views • 3 years ago
48:10Bay Area Vision Meeting: Unsupervised Feature Learning and Deep Learning
Google TechTalks
123K views • 15 years ago
41:07KEYNOTE / Robert Erdmann
PyCon US
24K views • 5 years ago
3 comments from Hacker News
By taking pictures of different angles a point cloud can be generated from the depth data which can be triangulated into a 3D model. I own a 3D printer and hope 3D scanning apps will soon be available. On the other hand, this demo is showing 3D generation by just filming the object: ▶ this video I wonder why this didn't take off.
I was at a local hackerspace and paid about $180 into a building class, but then it didn't work out and other things came up in my personal life and that of the person giving the classes. I think the $200-$300 range is where it's going to have to get to before 3d printers become wildly popular. Most people can overlook the low quality of the prints and the acetone wash if the cost of the printer is less than that of an iPad. Not everyone is going to want to use a 3d printer, but I feel like those of us who do want them but don't have them just don't feel that the price justifies the quality of current printers. The other barrier to ubiquitous 3d printer use is, in my opinion, a lack of cheap 3d scanning technology. Sure, you can use the laser line w/ turntable and camera scanner, but there are a number of issues, concavities and post-scan modeling work being the two largest I see. There are algorithms which can be used to take a webcam video feed and generate a 3d model from that, like proFORMA: ▶ this video ,but the source to that was never released. There are commercial products which can do this, but I think that until there is an open source/free option for this kind of scanning, it's going to be another barrier to people getting 3d printers.
It's rough but still pretty damn cool. Software that does this from footage can easily be in the thousands of dollars. Image the results if you pair a kinect with this: ▶ this video