#3DReconstruction

Kol Tregaskeskoltregaskes
2024-11-30

CAT4D: From Monocular Video to 4D Scene

CAT4D transforms monocular videos into multi-view 4D scenes for advanced 3D reconstruction. Leveraging generative priors and innovative multi-view diffusion, it improves scene consistency and fidelity.
arxiv.org/pdf/2411.18613

2024-11-17

Coming up next in the "#DH in Focus" series by #RosDH: Torsten Veit from @unigreifswald , who will explore how AI, 3D technologies, and computer vision are currently transforming manor house research and how they will shape its future.

πŸ—“οΈ Monday, Nov. 18, 2024, at 5:15 PM
πŸ“ #UniRostock & online
πŸ”— More details at: germanistik.uni-rostock.de/for

#digitalhistory #CulturalHeritage #BalticSea #HybridEvent #ComputerVision
#3DReconstruction #AI

Sebastian HageneuerArchaeoBasti@mastodon.online
2024-05-21

#500CharAbstracts: 🏺 @ArchaeoBasti & H. Franzmeier show, how to get from excavation data to a #3DReconstruction for an exhibition. They showcase the visualisation with the example of Pi-Ramesse in #Egypt, capital of Ramesses II. This case is especially interesting as the remains of the city are not visible any more. Special strategies were needed to create an aesthetically and educationally sound solution.
The article is available as #openaccess.

πŸ”— doi.org/10.11588/cipeg.2017.1.

This is a visualisation of Pi-Ramesse. Main view of the central island with canal and internal lake.
Sebastian HageneuerArchaeoBasti@mastodon.online
2024-03-19

#500CharAbstracts: I discuss how to present archaeological #3DReconstruction to a broader audience and show examples of the past and present. Beside the visually pleasing effects, #scicomm is often neglected.

This article is part of a broader book about communicating the ancient past, published in 2020 at Ubiquity: Communicating the Past in the Digital Age.

@ubiquity
#openaccess

πŸ”— doi.org/10.5334/bch.h

Reconstruction of Pi-Ramesses. Β© artefacts-berlin.de
2023-09-28
Testing #3DReconstruction with Neural Radiance Fields #NeRF. Recorded a video, extracting frames, estimating camera poses and then 40min of GPU computation results in this 5 000 000 points cloud, which looks really good even on the more subtle structures as for example the plant on the right.

It's quite simple with #nerfstudio, although it doesn't hurt if you kind of know what you are doing with your computer system and your GPU. I used a NVIDIA #RTX3070 with 8GB VRAM for training.
Dr Alan R.T. SpencerAlanRTSpencer@ecoevo.social
2022-11-22

#introduction time!  Well this is an interesting new place to be!

I’m Alan Spencer a Senior Teaching Fellow based at Imperial College, into #paleobotany and specialising in #3dreconstruction. Background in #geology and #archeology.

Interests are: research into 3D reconstruction of fossils. Mostly early seed-ferns, cycads, conifers, & earlier fungi-plant interactions. Augmented reality & 3D virtual classrooms, inclusivity on fieldwork, etc...

Now to figure out how #mastadon works!

Dmytro Mishkin πŸ‡ΊπŸ‡¦ducha_aiki@sigmoid.social
2022-11-22

Efficient Second-Order Plane Adjustment

Lipu ZHou

tl;dr: bundle adjustment, but for planes.

arxiv.org/abs/2211.11542.pdf

#computervision #arxiv #3dreconstruction

Berenike Reberenike@toot.io
2022-11-16

#introduction Hi! From today till Friday I take part in a colloquium of the DFK Paris. One of the best parts at the moment? My 3D model & reconstruction from my master thesis from 2021. #3dreconstruction #arthistory #digitalhumanities

Sebastian HageneuerArchaeoBasti@mastodon.online
2022-05-17

#500CharAbstracts: @ArchaeoBasti & H. Franzmeier show, how to get from excavation data to a #3DReconstruction for an exhibition. They showcase the visualisation with the example of Pi-Ramesse in #Egypt, capital of Ramesses II. This case is especially interesting as the remains of the city are not visible any more. Special strategies were needed to create an aesthetically and educationally sound solution.
The article is available as #openaccess.

πŸ”— doi.org/10.11588/cipeg.2017.1.

This is a visualisation of Pi-Ramesse. Main view of the central island with canal and internal lake.

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst