[ICCV-23] Paper List - 3D Generation-related
ICCV-23 paper list
目录
Oral Papers
3D from multi-view and sensors
Generative AI
Poster Papers
3D Generation (Neural generative models)
3D from a single image and shape-from-x
3D Editing
Face and gestures
Stylization
Dataset
Oral Papers
3D from multi-view and sensors
- Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields
- Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural Radiance Fields
- LERF: Language Embedded Radiance Fields
- Mixed Neural Voxels for Fast Multi-view Video Synthesis
- Multi-Modal Neural Radiance Field for Monocular Dense SLAM with a Light-Weight ToF Sensor
- Diffusion-Guided Reconstruction of Everyday Hand-Object Interaction Clips
- Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions
- Neural Haircut: Prior-Guided Strand-Based Hair Reconstruction
- ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes
- EgoLoc: Revisiting 3D Object Localization from Egocentric Videos with Visual Queries
- Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields (pdf)：Instant-NGP提出grid-based NeRF，对NeRF提速。Grid-based NeRF通常存在锯齿问题（alias），因此mip-NeRF 360提出将采样由射线变为圆锥，实现抗锯齿（Anti-aliasing）。但是，mip-NeRF 360不能很好的与Instant-NGP结合，因此本文提出了zip-NeRF。
- LERF: Language Embedded Radiance Fields (pdf) ：LERF是DFF的后续工作，在DINO features的基础上，额外引入了CLIP features，帮助实现NeRF中的细粒度定位和分类。
Generative AI
- TexFusion: Synthesizing 3D Textures with Text-Guided Image Diffusion Models
- Generative Novel View Synthesis with 3D-Aware Diffusion Models
- VQ3D: Learning a 3D-Aware Generative Model on ImageNet
Poster Papers
3D Generation (Neural generative models)
- GRAM-HD: 3D-Consistent Image Generation at High Resolution with Generative Radiance Manifolds
- Generative Multiplane Neural Radiance for 3D-Aware Image Generation
- Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model Using Pixel-Aligned Reconstruction Priors
- Towards High-Fidelity Text-Guided 3D Face Generation and Manipulation Using only Images
- ATT3D: Amortized Text-to-3D Object Synthesis
- Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation
- GETAvatar: Generative Textured Meshes for Animatable Human Avatars
- Mimic3D: Thriving 3D-Aware GANs via 3D-to-2D Imitation
- DreamBooth3D: Subject-Driven Text-to-3D Generation
- 3D-aware Image Generation using 2D Diffusion Models
- Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and Reconstruction
3D from a single image and shape-from-x
- Accurate 3D Face Reconstruction with Facial Component Tokens
- HiFace: High-Fidelity 3D Face Reconstruction by Learning Static and Dynamic Details
- Zero-1-to-3: Zero-shot One Image to 3D Object
- Deformable Model-Driven Neural Rendering for High-Fidelity 3D Reconstruction of Human Heads Under Low-View Settings
3D Editing
- Vox-E: Text-Guided Voxel Editing of 3D Objects
- FaceCLIPNeRF: Text-driven 3D Face Manipulation using Deformable Neural Radiance Fields
- SKED: Sketch-guided Text-based 3D Editing
- Seal-3D: Interactive Pixel-Level Editing for Neural Radiance Fields
Face and gestures
- Speech4Mesh: Speech-Assisted Monocular 3D Facial Reconstruction for Speech-Driven 3D Facial Animation
- Imitator: Personalized Speech-driven 3D Facial Animation
- EmoTalk: Speech-Driven Emotional Disentanglement for 3D Face Animation
- SPACE: Speech-driven Portrait Animation with Controllable Expression
Stylization
- Diffusion in Style
- Creative Birds: Self-Supervised Single-View 3D Style Transfer
- StyleDomain: Efficient and Lightweight Parameterizations of StyleGAN for One-shot and Few-shot Domain Adaptation
- StylerDALLE: Language-Guided Style Transfer Using a Vector-Quantized Tokenizer of a Large-Scale Generative Model
- X-Mesh: Towards Fast and Accurate Text-driven 3D Stylization via Dynamic Textual Guidance
- Locally Stylized Neural Radiance Fields
- DS-Fusion: Artistic Typography via Discriminated and Stylized Diffusion
- Multi-Directional Subspace Editing in Style-Space
- StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion Models
- All-to-Key Attention for Arbitrary Style Transfer
- DeformToon3D: Deformable Neural Radiance Fields for 3D Toonification
- Anti-DreamBooth: Protecting Users from Personalized Text-to-image Synthesis
- Neural Collage Transfer: Artistic Reconstruction via Material Manipulation
Dataset
- H3WB: Human3.6M 3D WholeBody Dataset and Benchmark
- SynBody: Synthetic Dataset with Layered Human Models for 3D Human Perception and Modeling
- Human-centric Scene Understanding for 3D Large-scale Scenario