Results 61 to 70 of about 8,361,437 (204)

Efficient View Synthesis with Neural Radiance Distribution Field [PDF]

, 2023
Recent work on Neural Radiance Fields (NeRF) has demonstrated significant advances in high-quality view synthesis. A major limitation of NeRF is its low rendering efficiency due to the need for multiple network forwardings to render a single pixel ...
Wang, Jinglu, Lu, Yan, Han, Xiaoguang, Wu, Yushuang, Li, Xiao, Cui, Shuguang   +5 more
core   +1 more source

IW-NeRF: Using Implicit Watermarks to Protect the Copyright of Neural Radiation Fields

Applied Sciences
The neural radiance field (NeRF) has demonstrated significant advancements in computer vision. However, the training process for NeRF models necessitates extensive computational resources and ample training data.
Lifeng Chen, Chaoyue Song, Jia Liu, Wenquan Sun, Weina Dong, Fuqiang Di   +5 more
doaj   +1 more source

NePO: Neural Point Octrees for Large‐Scale Novel View Synthesis

Computer Graphics Forum, EarlyView.
We introduce Neural Point Octrees (NePOs), a scalable radiance field representation that organises point clouds hierarchically for efficient optimisation and rendering of large scale scenes. NePOs enable level of detail selection, joint refinement of appearance and camera poses, and real‐time rendering of hundreds of millions of points.
Noah Lewis, Darius Rückert, Marc Stamminger, Linus Franke   +3 more
wiley   +1 more source

Hyperspectral Neural Radiance Field Method Based on Reference Spectrum

IEEE Access
The Neural Radiance Field (NeRF) method for datasets is gaining attention for its wide applications and research value. Hyperspectral images have also gained many applications in recent years.
Runchuan Ma, Sailing He
doaj   +1 more source

A Survey of Methods for Constructing 3D Urban Models From Point Clouds

Computer Graphics Forum, EarlyView.
The survey outlines a general process for constructing 3D models from point clouds and categorizes the methods based on their use of templates, basic surface primitives, hybrid approaches, or linear primitives. Additionally, the survey reviews the datasets and benchmarks that are essential for testing and training the developed methods.
Chiara Romanengo, Silvia Biasotti, Bianca Falcidieno, Daniela Cabiddu, Elia Moscoso Thompson, Michela Spagnuolo   +5 more
wiley   +1 more source

Improving 3D Object Detection in Neural Radiance Fields With Channel Attention

CAAI Transactions on Intelligence Technology
In recent years, 3D object detection using neural radiance fields (NeRF) has advanced significantly, yet challenges remain in effectively utilising the density field.
Minling Zhu, Yadong Gong, Dongbing Gu, Chunwei Tian   +3 more
doaj   +1 more source

RotGS: Rotation‐Guided 3D Gaussian Splatting for Turntable Sequences without Structure‐from‐Motion

Computer Graphics Forum, EarlyView.
Abstract The field of 3D reconstruction from multi‐view images has advanced rapidly thanks to 3D Gaussian Splatting (3DGS), which enables efficient and photorealistic scene representation. However, optimizing 3DGS requires high‐quality images from various viewpoints with accurate camera poses.
Kyumin Kim, Dohae Lee, Hanul Baek, In‐Kwon Lee   +3 more
wiley   +1 more source

Point-NeRF: Point-based Neural Radiance Fields

, 2023
Volumetric neural rendering methods like NeRF generate high-quality view synthesis results but are optimized per-scene leading to prohibitive reconstruction time.
Xu, Zexiang, Philip, Julien, Bi, Sai, Neumann, Ulrich, Xu, Qiangeng, Sunkavalli, Kalyan, Shu, Zhixin   +6 more
core  

3D Character Reconstruction from Hand‐drawn Model Sheets

Computer Graphics Forum, EarlyView.
Abstract Hand‐drawn model sheets are widely used in character design to define 3D shape and appearance through sparse multi‐view drawings. Reconstructing 3D characters from such sparse inputs has traditionally been challenging due to insufficient visual information.
Hyejeong Yoon, Wonjong Jang, Yoonha Hwang, Seungyong Lee   +3 more
wiley   +1 more source

Multi‐Spectral Gaussian Splatting with Neural Color Representation

Computer Graphics Forum, EarlyView.
Abstract 3D Gaussian Splatting (3DGS) [KKLD23] has transformed novel‐view synthesis from RGB images, yet remains restricted to the visible spectrum. Many applications, including agricultural monitoring, rely on multi‐spectral imaging, where spectral camera alignment and scalability pose major challenges.
Lukas Meyer, Josef Grün, Maximilian Weiherer, Bernhard Egger, Marc Stamminger, Linus Franke   +5 more
wiley   +1 more source