3D 재구축 / 디지털 트윈

A client hands you a single, cluttered photo of their living room and asks what it would look like r

@wildmindai

LEGO-SLAM: Language-Embedded Gaussian Optimization SLAM

@rsasaki0109

🐠🐟💙 Workflow

@snowscoop.studio

엔비디아 연구진이 텍스트와 물리적 제약 조건을 통해 고품질 3D 모션을 생성하는 모션 디퓨전 모델 'Kimodo'를 공개했습니다.

@choi.openai

📍 '수정 가능한' 데이터라니…

@miadstudio

-

@snowscoop.studio

DepthViz 2.0 출시

@orcawalk

구글이 150여 개국 260만 건의 역사적 데이터를 학습한 AI 기반 돌발 홍수 예측 모델 그라운드소스를 공개했습니다. 

@modulabs_

2025년에 마무리한 KIA_PV5 프로젝트의 1970~1980년대 배경 작업 과정을 공유합니다.

@h_steel_fbx

이미지 하나를 넣으면 3D 모델 생성 → 토폴로지 → UV → 텍스처 → 리깅 → 애니메이션까지 이어지는 전체 제작 과정을 한 워크플로 안에서 처리하는 AI 작업 공간입니다.

@choi.openai

뭔가 많이 아쉬운 두 번째 렌더

@hdn_silver

예를 들어 캐릭터 이미지나 제품 사진을 업로드하면, 별도의 3D 모델링 작업 없이 바로 출력 가능한 3D 모델로 변환됩니다.

@choi.openai

Seedance를 활용하면 노드 기반 시스템으로 각각의 단계를 빠르게 점검하고 반복 수정하는 과정이 매우 간편해집니다. 별도의 소프트웨어 이동 없이 한 공간에서 2D→3D 변환,

@ai.ivn

기존에는 패브릭 주름이나 작은 디테일이 다소 뭉개지는 느낌이 있었는데,

@yoon9718

3D 생성 플랫폼 Meshy에서 텍스트만 입력하면 완벽하게 이어지는(Seamless) PBR 텍스처를 만들어내는 기능을 공개했습니다.

@choi.openai

Tripo가 Tripo P1.0을 공개하며 Tripo Studio에 “Smart Mesh” 기능을 추가했습니다. 프롬프트만 입력하면 약 2초 만에 구조가 정리된 3D 메쉬를 생성하

@choi.openai

이번 해커톤은 AR 트랙과 VR 트랙으로 나뉘어 진행됩니다. 각 트랙 우승자에게는 Santa Monica College 혹은 박물관/미술관에서 작업을 선보일 수 있는 기회를 제공하

@pooper_in_la

check the link in bio to finally learn Houdini 🫡  

@nickmedukha

3D Artist and Illustrator Chris Jones has showcased a flexible human hand rig and shared some detail

@eighty_level

Have a great week friends🤩  

@marterium

gm, probably my most popular work 🕹️  

@synccreation

Meta Quest 3 → room scan → Pascal editor (built with

@j_brissonneau

Li and Zhu et al., "DROID-SLAM in the Wild"

@kwangmoo_yi

Wild action, casual capture, no problem.

@MoyangLi00

Gaussian splatting as part of the UI – a lot went into making this as performant as possible – for e

@Haxiomic

DROID-SLAM in the Wild

@zhenjun_zhao

converts single 2D images into interactable 3D models; with MonoArt.

@wildmindai

lukasHoel/video_to_world

@lukasHoel

Headed to Vancouver for #3DV to present 3 papers!

@ethanjohnweber

Leveling3D: Leveling Up 3D Reconstruction with Feed-Forward 3D Gaussian Splatting and Geometry-Aware

@zhenjun_zhao

ChenYutongTHU/GGPT

@ChenYutongTHU

rerun-io/examples-monorepo

@rerun-io

Started bringing a smart home idea I generated with Nano Banana 2 to life

@DilumSanjaya

EgoEdit

@RunjiaLi

cartographer-project/cartographer

@cartographer-project

dimensionalOS/dimos

@dimensionalOS

Videos are continuous projections of 3D worlds. After training on massive video data, does 3D unders

@zixuan_huang

3D wiping effect using MeshBVH #threejs

@auxdesigner

Not long ago, we thought this might be impossible. But here we are: our Gaussian Splatting videos ar

@gracia_vr

Mind blown by this new CVPR 2026 paper: tttLRM

@KIRI_Engine_App

Everyone is scaling VLAs with more robot data.

@robotsdigest

Spatial-TTT dropped!

@fangfu0830

Cheng et al., "ReCoSplat: Autoregressive Feed-Forward Gaussian Splatting via Render-and-Compare"

@kwangmoo_yi

Junyi42/LoGeR

@Junyi42

¡AGENTES INMOBILIARIOS, ESTO VA A VOLARLES LA CABEZA!

@ErickSky

Topo-Or [

@tolga_birdal

A single image can transform into a massive world composed of millions of splats with Marble. Formin

@theworldlabs

안녕하세요. 3D 에셋을 구하는 데에 어려움을 겪으시는 개발자분들이 계신 것 같아서, 제가 아는 무료 에셋 사이트를 타래로 공유해 보려고 합니다.

@WTESgame

Excited to share that our paper LASER has been accepted to #CVPR2026!

@TianyeJerryDing

Visionary-Laboratory/holi-spatial

@Visionary-Laboratory

Spherical-GOF: Geometry-Aware Panoramic Gaussian Opacity Fields for 3D Scene Reconstruction

@janusch_patas

Announcing "Holi-Spatial"!

@10027lifuguan

boschresearch/3dgeer

@boschresearch

PixARMesh reconstructs full indoor scenes from a single photo directly into lightweight meshes.

@wildmindai

"Track4World: Feedforward World‑centric Dense 3D Tracking of All Pixels"

@Almorgand

Over the weekend, I sent Droid on a Mission to create a home automation app. I wanted it to look lik

@EnoReyes

I made a procedural anime cloud generator for Blender

@Mix3Design

𝗢𝗻𝗲 𝗺𝗲𝗺𝗼𝗿𝘆 𝗰𝗮𝗻’𝘁 𝗿𝘂𝗹𝗲 𝘁𝗵𝗲𝗺 𝗮𝗹𝗹.

@junyi42

Introducing OnlineSI: Taming Large Language Model for Online 3D Understanding and Grounding!

@ZixianLiu03

Recreated this in

@shawn_kel

spla-tam/SplaTAM

@spla-tam

ポートフォリオサイト一旦デプロイ！

@KCDCjpn

Crea un modelo 3D de una habitación en menos de 2 minutos.

@MiguelMaestroIA

#ICLR2026

@ChuanxiaZ

huggingface/meshgen

@huggingface

Fantasy-AMAP/fantasy-world

@Fantasy-AMAP

Do Visual Imaginations Improve Vision-and-Language Navigation Agents?

@rsasaki0109

ZipMap: Linear-Time Stateful 3D Reconstruction with Test-Time Training

@zhenjun_zhao

inspatio/worldfm

@inspatio

Blender render → 4D Gaussian Splatting training → Houdini

@emilsvfx

Stop searching Zillow for the perfect house.

@mhdfaran

lifuguan/IGGT_official

@lifuguan

HengyiWang/amb3r

@HengyiWang

Generated worlds can now be embedded anywhere!

@tamrrat

jingwu2121/reflect3r

@jingwu2121

finally experimenting with gaussian splats + Apple's ML SHARP model

@tjcages

Built DepthShift — takes a flat video + depth map and reprojects it into a live 3D particle field. V

@verymehari

threedle/ll3m

@threedle

Built my own Guassian Splat viewer for my app Interior AI

@levelsio

Can we talk about how insane Gemini 3.1 Pro is at webgl

@MengTo

Testing out Apple’s SHARP to generate a point cloud from a 2D Image then using

@SinzVII

This started as a flat 2D floor plan.

@techhalla

It started as an experiment in

@arturitu

weihan1/growflow

@weihan1

NVlabs/sage

@NVlabs

It started as ThreeJS 3d character demo...

@chongdashu

Finished the main part of the infinite terrain I had in mind

@mesqme

jaskirat1616/Apple-ml-sharp-rerun

@jaskirat1616

Doing some preparations for the Steam page! Started reworking the 3D models.

@holypersondev

apple/ARKitScenes

@apple

@threejs

facebookresearch/ShapeR

@facebookresearch

ywyue/RoomFormer

@ywyue

maturk/dn-splatter

@maturk

CloudAI-X/threejs-skills

@CloudAI-X

MrNeRF/LichtFeld-Studio

@MrNeRF

DepthAnything/PromptDA

@DepthAnything

XuqianRen/AGS_Mesh

@XuqianRen

ColmapView/Colmapview.github.io

@ColmapView

rerun-io/eye_control_example

@rerun-io

Introducing ShapeR, a method for robust conditional 3D shape generation from casually captured seque

@yawarnihal

Our isometric websites are always

@Praha37v

Vibe coded something new today

@thebuggeddev

SAM changed the game for 2D image segmentation.

@HuaizuJiang

zbw001/TAPIP3D

@zbw001

henry123-boy/SpaTrackerV2

@henry123-boy

Intermission switch. #b3d

@reijowrites

Go back. #b3d

@reijowrites

teslamotors/react-native-camera-kit

@teslamotors

xiangechen/chili3d

@xiangechen

usnistgov/SFA

@usnistgov

Tamagotchi

@PeterTarka

chieaid24/PMI-Auto-Generator-Desc

@chieaid24

thingraph/bim-viewer

@thingraph

thingraph/dwg-viewer-example

@thingraph

this is stunning..

@Aurelien_Gz

Satisfaction. #b3d

@reijowrites

Excited to share our latest work, WildGS-SLAM: Monocular Gaussian Splatting SLAM in Dynamic Environm

@jianhao75895505

zerchen/hort

@zerchen

facebookresearch/fast3r

@facebookresearch

mega-sam/mega-sam

@mega-sam

DavidHDev/react-bits

@DavidHDev

princeton-vl/infinigen

@princeton-vl

kovacsv/occt-import-js

@kovacsv

concept-graphs/concept-graphs

@concept-graphs

margelo/react-native-filament

@margelo

From Discourse: "How to setup objs for select & move"

@0xca0a

callstack/react-native-visionos

@callstack

xxlong0/Wonder3D

@xxlong0

graphdeco-inria/gaussian-splatting

@graphdeco-inria

MoyGcc/vid2avatar

@MoyGcc

satoshi0212/visionOS_30Days

@satoshi0212

FarazzShaikh/THREE-CustomShaderMaterial

@FarazzShaikh

CMU-Perceptual-Computing-Lab/openpose

@CMU-Perceptual-Computing-Lab

apple/ml-neuman

@apple

homuler/MediaPipeUnityPlugin

@homuler

godotengine/godot

@godotengine

google/model-viewer

@google

mrdoob/three.js

@mrdoob

pmndrs/drei

@pmndrs

pmndrs/gltfjsx

@pmndrs

pmndrs/react-three-fiber

@pmndrs