Angela Dai
Associate Professor
Works on 3D scene understanding and reconstruction from commodity RGB-D data, using generative 3D deep learning to create semantically grounded 3D environments.
Angela Dai's lab at TUM works on 3D vision & neural rendering, recognition, detection & segmentation and generative 3D & visual content. 45 papers since 2023, 44 at top venues, mostly CVPR and ICCV. 7 PhD students.
At a glance
- papers, 2023–2026
- 12/ year
- at T1 venues
- 98%
- PhD students
- 7
- MS students
- 0
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 11 papers2026 · T1 11 · collecting
- 12 papers2025 · T1 12
- 13 papers2024 · T1 12
- 9 papers2023 · T1 9
T1 venuesOtherVenues: CVPR, ICCV, NeurIPS, ECCV, ICLR
Research focuspaper titles that use each topic's words
- 3D vision & neural rendering7 papers
- Generative 3D & visual content3 papers
- Recognition, detection & segmentation2 papers
- scene understanding2 papers
- grounded0 papers
Paper list
From the lab homepage.
202611 papers
- ICLR (T1)Animating the Uncaptured: Humanoid Mesh Animation with Video Diffusion Models
- NeurIPS (T1)Diff3R: Feed-forward 3D Gaussian Splatting with Uncertainty-aware Differentiable Optimization
- NeurIPS (T1)GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction
- ICML (T1)HOI-PAGE: Zero-Shot Human-Object Interaction Generation with Part Affordance Guidance
- NeurIPS (T1)Lookalike3D: Seeing Double in 3D
- CoRL (T1)MotionDisco: Motion Discovery for Extreme Humanoid Loco-Manipulation
- ECCV (T1)Seen2Scene: Completing Realistic 3D Scenes with Visibility-Guided Flow
- ECCV (T1)TriFlow: Generating Artist-Like 3D Mesh Topology via Nearest-Vertex Vector Fields
- NeurIPS (T1)UnfoldArt: Zero-Shot Recovery of Full Articulated 3D Objects from Text or Image
- ECCV (T1)WorldAgents: Can Foundation Image Models be Agents for 3D World Models?
- ECCV (T1)WorldMesh: Generating Navigable Multi-Room 3D Scenes via Mesh-Conditioned Image Diffusion
202512 papers
- ICCV (T1)DiffuMatch: Category-Agnostic Spectral Diffusion Priors for Robust Non-rigid Shape Matching
- CVPR (T1)DNF: Unconditional 4D Generation with Dictionary-based Neural Fields
- ICCV (T1)ExCap3D: Expressive 3D Scene Understanding via Object Captioning with Varying Detail
- ICCV (T1)GaussianSpeech: Audio-Driven Personalized 3D Gaussian Avatars
- CVPR (T1)LT3SD: Latent Trees for 3D Scene Diffusion
- CVPR (T1)MeshArt: Generating Articulated Meshes with Structure-guided Transformers
- ICCV (T1)MeshPad: Interactive Sketch-Conditioned Artist-Designed Mesh Generation and Editing
- ICLR (T1)Open-YOLO 3D: Towards Fast and Accurate Open-Vocabulary 3D Instance Segmentation
- CVPR (T1)PrEditor3D: Fast and Precise 3D Shape Editing
- ICCV (T1)QuickSplat: Fast 3D Surface Reconstruction via Learned Gaussian Initialization
- ICCV (T1)ScanEdit: Hierarchically-Guided Functional 3D Scan Editing
- CVPR (T1)SceneFactor: Factored Latent 3D Diffusion for Controllable 3D Scene Generation
202413 papers
- CVPR (T1)CG-HOI: Contact-Guided 3D Human-Object Interaction Generation
- NeurIPS (T1)Coherent 3D Scene Diffusion From a Single RGB Image
- TOG (T1)DiffCAD: Weakly-Supervised Probabilistic CAD Model Retrieval and Alignment from an RGB Image
- CVPR (T1)DiffuScene: Denoising Diffusion Models for Generative Indoor Scene Synthesis
- CVPR (T1)DPHMs: Diffusion Parametric Head Models for Depth-based Tracking
- NeurIPS Datasets and BenchmarksDrivAerNet++: A Large-Scale Multimodal Car Dataset with Computational Fluid Dynamics Simulations and Deep Learning Benchmarks
- CVPR (T1)FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models
- CVPR (T1)FutureHuman3D: Forecasting Complex Long-Term 3D Human Behavior from Video Observations
- CVPR (T1)GenZI: Zero-Shot 3D Human-Scene Interaction Generation
- SIGGRAPH Asia (T1)L3DG: Latent 3D Gaussian Diffusion
- CVPR (T1)MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers
- CVPR (T1)PaSCo: Urban 3D Panoptic Scene Completion with Uncertainty Awareness
20239 papers
- SIGGRAPH (T1)ClipFace: Text-guided Editing of Textured 3D Morphable Models
- ICCV (T1)HyperDiffusion: Generating Implicit Neural Fields with Weight-Space Diffusion
- CVPR (T1)Learning 3D Scene Priors with 2D Supervision
- CVPR (T1)Mask3D: Pre-training 2D Vision Transformers by Learning Masked 3D Priors
- ICCV (T1)Mesh2Tex: Generating Mesh Textures from Image Queries
- CVPR (T1)Neural Part Priors: Learning to Optimize Part-Based Object Completion in RGB-D Scans
- CVPR (T1)ObjectMatch: Robust Registration using Canonical Object Correspondences
- CVPR (T1)Panoptic Lifting for 3D Scene Understanding with Neural Fields
- ICCV (T1)ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes
Where alumni went
Alumni are not listed on the lab homepage.
Apply to TUMRolling
Not announced yet
Computer Science Doctorate (CIT)
- Round
- Rolling (position-based)
- Opens
- —
- GRE
- See program page
- English
- Very good written and spoken English; no minimum test score
- Fee
- See program page
Confirmed on the program page.
Similar labsnearest by research summary
- USUW
Brian Curless
- 3D vision & neural rendering
- CNPeking
Gang Zeng
- 3D vision & neural rendering
- CNSJTU
Qiang Hu
- 3D vision & neural rendering
- USUMD
Jia-Bin Huang
- 3D vision & neural rendering
- CHETH Zurich
Siyu Tang
- 3D vision & neural rendering
- GBUCL
Lourdes Agapito
- 3D vision & neural rendering
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.