pengc02/my-paper-list

my paper reading list

1. Network Structure

• StyleGan系列：

1. A Style-Based Generator Architecture for Generative Adversarial Networks
   • NVlabs/stylegan: StyleGAN - Official TensorFlow Implementation (github.com)
2. Analyzing and Improving the Image Quality of StyleGAN
   • NVlabs/stylegan2: StyleGAN2 - Official TensorFlow Implementation (github.com)

2. Avatar

2.1. Face

• A Morphable Model For The Synthesis Of 3D Faces

  • https://github.com/MichaelMure/3DMM

• FLAME: Learning a model of facial shape and expression from 4D scans

  • FLAME (mpg.de)

3. 3D Generation

• DreamFusion

  • DreamFusion: Text-to-3D using 2D Diffusion (dreamfusion3d.github.io)
  • Description：propose SDS loss

• ProlificDreamer

  • ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillation (tsinghua.edu.cn)
  • Description：propose VDS loss

• Magic3d

  • Magic3D: High-Resolution Text-to-3D Content Creation (nvidia.com)
  • 两阶段生成，coarse-to-fine

• Magic123

  • Magic123: One Image to High-Quality 3D Object Generation Using Both 2D and 3D Diffusion Priors (guochengqian.github.io)

• DreamCraft3D

  • DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior (mrtornado24.github.io)

• DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation

  • DreamGaussian

3D大模型

• zero-123

  • provide objaverse dataset Objaverse (allenai.org)
  • train a novel-view generator

• zero-123++

• One-2-3-45

  • One-2-3-45
  • Use Zero123 to generate multi-view images for the input view, and then aim to lift them up to 3D space

• LRM: Large Reconstruction Model for Single Image to 3D

  • https://scalei3d.github.io/LRM/

• Instant3D: Fast Text-to-3D with Sparse-view Generation and Large Reconstruction Model

  • https://instant-3d.github.io

• DMV3D: Denoising Multi-view Diffusion Using 3D Large Reconstruction Model

  • https://dmv3d.github.io

• PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Prediction

  • https://totoro97.github.io/pf-lrm

4. 3D-Representation

• Indoor Segmentation and Support Inference from RGBD Images

  • 相当早期的标注深度数据集
  • 提出深度和normal预测方法
  • ECCV-12

• GeoNet

  • xjqi/GeoNet: GeoNet: Geometric Neural Network for Joint Depth and Surface Normal Estimation (github.com)
  • CVPR 2018

• NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction

  • NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction (lingjie0206.github.io)
  • NeurIPS 2021 (Spotlight)
  • signed distance function (SDF) 与 neural SDF 表示

• 3D Gaussian Splatting for Real-Time Radiance Field Rendering

  • 3D Gaussian Splatting for Real-Time Radiance Field Rendering (inria.fr)
  • TOG 2023
  • 提供了一种整体**类似Nerf但是效率更高的高斯点云方法

To be done

MVDream: Multi-view Diffusion for 3D Generation https://mv-dream.github.io/

adobe灌了四篇：

LRM: Large Reconstruction Model for Single Image to 3D https://scalei3d.github.io/LRM/

Instant3D: Fast Text-to-3D with Sparse-view Generation and Large Reconstruction Model https://instant-3d.github.io

DMV3D: Denoising Multi-view Diffusion Using 3D Large Reconstruction Model https://dmv3d.github.io

PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Prediction https://totoro97.github.io/pf-lrm

SweetDreamer: Aligning Geometric Priors in 2D diffusion for Consistent Text-to-3D

DMTet Deep Marching Tetrahedra: a Hybrid Representation for High-Resolution 3D Shape Synthesis (nvidia.com)

EG3D EG3D: Efficient Geometry-aware 3D GANs (nvlabs.github.io)

One-2-3-45++ (sudo-ai-3d.github.io)

Instant Neural Graphics Primitives with a Multiresolution Hash Encoding (nvlabs.github.io) instant-ngp

justimyhxu/awesome-3D-generation: A curated list of awesome 3d generation papers (github.com)