MoSE3: Learning World-Space SE(3) at Every Pixel
https://arxiv.org/abs/2610.03716
Surflo Consistent 3D Surface Flow Model with Global State
https://anttwo.github.io/surflo/
Déjà View: Looping Transformers for Multi-View 3D Reconstruction
https://github.com/nv-tlabs/dvlt
Real-time 3D full-body reconstruction
https://github.com/AmmarkoV/SAM3DBody-cpp
Pixal3D: Pixel-Aligned 3D Generation from Images
https://github.com/TencentARC/Pixal3D
Modly - Local, open source, AI-powered image-to-3D mesh generatio
https://github.com/lightningpixel/modly
Boxer: Robust Lifting of Open-World 2D Bounding Boxes to 3D
https://github.com/facebookresearch/boxer
Free-Range Gaussians: Non-Grid-Aligned Generative 3D Gaussian Reconstruction
https://arxiv.org/abs/2604.04874
SpatialLM: Training Large Language Models for Structured Indoor Modeling
https://github.com/manycore-research/SpatialLM
SAM 3D Body: Robust Full-Body Human Mesh Recovery
https://github.com/facebookresearch/sam-3d-body
SAM 3D Objects with Multi-view Images
https://github.com/devinli123/MV-SAM3D
Utonia: Toward One Encoder for All Point Clouds
https://github.com/Pointcept/Utonia
tttLRM: Test-Time Training for Long Context and Autoregressive 3D Reconstruction
https://cwchenwang.github.io/tttLRM/
PhysX-Anything: Simulation-Ready Physical 3D Assets from Single Image
https://physx-anything.github.io
Sharp Monocular View Synthesis in Less Than a Second
https://github.com/apple/ml-sharp
TRELLIS.2: Native and Compact Structured Latents for 3D
https://microsoft.github.io/TRELLIS.2/
MapAnything: Universal Feed-Forward Metric 3D Reconstruction
https://github.com/facebookresearch/map-anything
Native and Compact Structured Latents for 3D Generation
https://github.com/microsoft/TRELLIS.2
PhysX-Anything: Simulation-Ready Physical 3D Assets from Single Image
https://github.com/ziangcao0312/PhysX-Anything
Meta SAM 3D
https://ai.meta.com/sam3d/
Fast and Universal 3D reconstruction model for versatile tasks
https://github.com/Tencent-Hunyuan/HunyuanWorld-Mirror
https://github.com/Tencent-Hunyuan/Hunyuan3D-2.1
Rectified Point Flow: Generic Point Cloud Pose Estimation
https://rectified-pointflow.github.io/
FlashWorld: High-quality 3D Scene Generation within Seconds
https://github.com/imlixinyang/FlashWorld
OmniPart: Part-Aware 3D Generation with Semantic Decoupling and Structural Cohesion
https://github.com/HKU-MMLab/OmniPart
Steerable Scene Generation with Post Training and Inference-Time Search
https://steerable-scene-generation.github.io/
Open-YOLO 3D: Towards Fast and Accurate Open-Vocabulary 3D Instance Segmentation
https://github.com/aminebdj/OpenYOLO3D
UniGS: Unified Geometry-Aware Gaussian Splatting for Multimodal Rendering
https://arxiv.org/abs/2510.12174
SuperDec: 3D Scene Decomposition with
Superquadric Primitives
https://github.com/elisabettafedele/superdec
FlowR: Flowing from Sparse to Dense 3D Reconstructions
https://tobiasfshr.github.io/pub/flowr/
Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation
https://research.nvidia.com/labs/toronto-ai/lyra/
CityDreamer4D: Compositional Generative Model of Unbounded 4D Cities
https://github.com/hzxie/CityDreamer4D.
4DNeX: Feed-Forward 4D Generative Modeling Made Easy
https://4dnex.github.io/