single-view-3d

📚 CVPR 2024 - Single-view 3D Reconstruction Paper List

📄 Paper List

---

1. Splatter Image: Ultra-Fast Single-View 3D Reconstruction

Authors: Stanislaw Szymanowicz, Christian Rupprecht, Andrea Vedaldi
Affiliations: University of Oxford

Highlights:

• 提出了一种基于高斯散点的单图三维重建方法。
• 输入图像经神经网络映射为每像素 3D 高斯分布，实时重建速度高达 38 FPS。
• 高效、端到端，重建质量与速度远超现有方法。

🔗 Resources

📄 Paper PDF | 🔗 Project Page | 💻 Code Repository

🔍 Keywords: Single-view 3D Reconstruction, Gaussian Splatting, Real-time Rendering
🗂️ Dataset: Synthetic and real multi-category datasets

---

2. Triplane Meets Gaussian Splatting: Fast and Generalizable Single-View 3D Reconstruction with Transformers

Authors: Zi-Xin Zou, Zhipeng Yu, Yuan-Chen Guo, Yangguang Li, Ding Liang, Yan-Pei Cao, Song-Hai Zhang
Affiliations: Tsinghua University, SenseTime Research

Highlights:

• 结合 Triplane 表示与高斯散点，使用 Transformer 构建单图 3D 重建模型。
• 泛化能力强，在多数据集上保持一致高性能。

🔗 Resources

📄 Paper PDF

🔍 Keywords: Single-view 3D Reconstruction, Triplane Representation, Gaussian Splatting, Transformers
🗂️ Dataset: Various benchmark datasets

---

3. SiTH: Single-view Textured Human Reconstruction with Image-Conditioned Diffusion

Authors: Hsuan-I Ho, Jie Song, Otmar Hilliges
Affiliations: ETH Zurich

Highlights:

• 利用图像条件扩散模型，实现纹理化人体的单图重建。
• 生成完整 3D 人体仅需 2 分钟，结合条件控制与纹理生成。

🔗 Resources

📄 Paper PDF | 💻 Code Repository

🔍 Keywords: Single-view 3D Reconstruction, Human Reconstruction, Diffusion Models
🗂️ Dataset: Human3.6M, SURREAL

---

4. Slice3D: Multi-Slice Occlusion-Revealing Single View 3D Reconstruction

Authors: Yizhi Wang, Wallace Lira, Wenqi Wang, Ali Mahdavi-Amiri, Hao Zhang
Affiliations: Simon Fraser University

Highlights:

• 多切片重建方法能有效揭示遮挡区域，提高遮挡场景下的重建完整性。
• 切片结构能平衡精细建模与效率，提升遮挡理解能力。

🔗 Resources

📄 Paper PDF

🔍 Keywords: Single-view 3D Reconstruction, Occlusion Handling, Multi-Slice Representation
🗂️ Dataset: ShapeNet, Pix3D

---

5. Gamba: Marry Gaussian Splatting with Mamba for Single View 3D Reconstruction

Authors: Qiuhong Shen, Zike Wu, Xuanyu Yi, Pan Zhou, Hanwang Zhang, Shuicheng Yan, Xinchao Wang
Affiliations: National University of Singapore

Highlights:

• 将 Gaussian Splatting 与 Mamba 架构结合，实现超快单图 3D 重建。
• 推理时间仅需 0.05 秒，比优化方法快 1000 倍。

🔗 Resources

📄 Paper PDF | 🔗 Project Page

🔍 Keywords: Single-view 3D Reconstruction, Gaussian Splatting, Mamba Architecture
🗂️ Dataset: Objaverse, GSO

---

This site is open source. Improve this page.