#NeRF

This open-source project, initiated by Berkeley students, encourages community collaboration and provides valuable resources for both beginners and seasoned users. Explore modular and interpretable systems for creating, training, and testing NeRFs. Access tutorials for leveraging custom data, visualizing 3D scenes, and making efficient use of NeRF technology. Join the growing NeRF community to contribute to the project and explore advanced options like real-time training visualization and easy benchmarking.

This collection offers insights into the advancements in Radiance Fields, particularly NeRF and 3D Gaussian Splatting, serving as a comprehensive guide for mastering 3D editing techniques. Access key papers, relevant surveys, and the latest research developments, providing a thorough understanding of radiance field-based 3D editing. Connect with the research community through continuous updates and collaborative openings, simplifying your journey in this evolving discipline.

K-Planes offers an extensible radiance field model for a variety of static and dynamic datasets, applicable across different scene dimensions. This tool enables explicit modeling of radiance fields over space and time, improving the visualization of appearances in 3D spaces. The project's setup utilizes PyTorch within a conda environment, providing scripts for comprehensive training and evaluation. Integration with NerfAcc and NerfStudio libraries ensures an efficient training and visualization process. Delve into capabilities such as novel camera trajectory rendering and quality metric evaluation, beneficial for advancing neural rendering technologies.

Discover 2160 innovative papers and open-source projects presented at ICCV 2023. This collection covers a wide range of computer vision topics, including vision-language models, 3D object detection, and neural radiance fields. Perfect for researchers and enthusiasts aiming to keep abreast of the latest advancements without promotional exaggeration. Dive into past CV conference papers and engage with academic groups for in-depth discussions. Access domain-specific resources and uncover progressions in areas like self-supervised learning, image editing, and diffusion models.

Generalizable NeRF Transformer (GNT) employs transformers to reconstruct Neural Radiance Fields efficiently. It utilizes a dual-stage architecture with attention mechanisms for scene representation and rendering, achieving results comparable to state-of-the-art methods. GNT demonstrates improvements in PSNR and capabilities in depth and occlusion inference, optimized for various datasets.

The BAD-Gaussians project leverages the nerfstudio framework to implement Gaussian-based methods for effective deblurring and novel-view synthesis of motion-blurred images. This approach optimizes the Deblur-NeRF datasets by enhancing clarity and rendering quality through robust Gaussian splatting techniques. It supports both synthetic and real-world data processing with adjustable parameters for camera pose interpolation and image scaling, offering comprehensive instructions for installation, dataset setup, training, video rendering, and 3D Gaussian export. This makes it a valuable resource for researchers and developers in the fields of image processing and computer vision.

This project benchmarks cutting-edge unbounded Neural Radiance Fields (NeRF) algorithms, offering a streamlined, high-performance code repository. The results highlight comparisons with widely-used methods such as NeRF++, Plenoxels, and DVGO, showcasing notable PSNR improvements. With practical guidelines on installation, data processing, and training, this project is a valuable resource for researchers and developers aiming for optimized neural radiance field performance using public datasets. The project also provides ongoing updates and comprehensive documentation for building custom NeRFs.

InstantMesh provides an efficient framework for generating 3D meshes from single images, utilizing LRM/Instant3D architecture. Features include zero123++ fine-tuning, available model weights, and Gradio demos via Huggingface. Suitable for single and dual-GPU use, it also supports Docker. Recommended environment includes Python 3.10, PyTorch 2.1.0, and CUDA 12.1. It offers mesh generation with varied resolutions and command line tools for detailed modeling, with training code available for further model development.

This project delivers a PyTorch and Taichi-driven implementation of the instant-ngp NeRF training pipeline. It includes guides for installation, dataset processing, and training scripts applicable to both preprocessed datasets and custom videos. Compatibility with Linux and RTX graphics cards is emphasized, with half2 optimization available for improved performance. Mobile deployment is feasible through Taichi AOT, enabling real-time iOS rendering. The project also supports text-to-3D rendering as a backend for the stable-dreamfusion project. Explore streamlined procedures and comprehensive instructions for efficient NeRF training and deployment.

This repository provides an organized collection of research papers, code, and resources related to GANs and NeRFs, with a focus on audio and image-driven talking head synthesis. It includes a variety of datasets, tools, and metrics, making it a useful resource for those researching or developing realistic digital human faces. The platform is regularly updated with community contributions, ensuring its status as a continually evolving resource. Many papers are accessible through 'arXiv' and other academic channels, although some require an academic license.

This repository provides a comprehensive collection of research and resources on Implicit Representations and Neural Radiance Fields (NeRFs) in the context of Robotics and Reinforcement Learning (RL). It features a curated selection of key papers, relevant workshops, and progress updates on integrating NeRFs into areas such as object reconstruction, pose estimation, SLAM, and manipulation. Regular updates make it a central resource for researchers interested in exploring these technologies. Learn about significant features, recent developments, and practical applications illustrating the impact of NeRFs on robotic systems.

A robust framework for X-ray 3D reconstruction, featuring novel view synthesis and CT methods across 11 cutting-edge approaches, including NeRF-based techniques. Facilitates research with tools for visualization, data generation, and access to pre-trained models and comprehensive datasets, optimizing accuracy in sparse-view scenarios.

This project provides a comprehensive set of nodes for efficient 3D asset generation in ComfyUI, using algorithms like 3DGS, NeRF, and InstantMesh. It handles inputs including Mesh and UV Texture, and supports major operating systems and GPU setups. The ComfyUI-Manager facilitates easy installation of pre-built components, with options for automated or semi-automated builds. A diverse range of models enables conversion of single images into detailed 3D meshes with texture. Features include tools like StableFast3D and CharacterGen, and advanced functions such as 3D Gaussian Splatting and Instant NGP. Supported by a development community and workflow guides, it serves as a versatile tool for today's 3D creators.

This article examines the implementation of Kolmogorov-Arnold Networks (KAN) with a variety of basis functions, including B-Spline, Fourier, and Gaussian RBF, applied to the NeRF equation. Utilizing the nerfstudio framework, the project aims to advance the comprehension and application of neural radiance fields. It provides a detailed comparison of KAN model performance on an RTX-3090, highlighting metrics like speed, PSNR, and SSIM. The article is open to feedback for improvements and provides clear installation and usage instructions.

This project offers a detailed curation of Text-to-3D and Diffusion-to-3D methods inspired by awesome-NeRF, featuring numerous research papers on converting text to 3D models using techniques like Neural Radiance Fields (NeRF) and advanced diffusion models. It is regularly updated with new resources including project page links, video tutorials, and cited literature with code repositories, providing a valuable resource for exploring innovative 3D synthesis developments. Applications range from object generation to novel view synthesis, serving as an informative guide for researchers and enthusiasts.

GeneFace is an audio-driven 3D face synthesis tool that excels in lip synchronization and expressiveness, even with out-of-domain audio. Utilizing a RAD-NeRF-based renderer, it supports real-time inference and cuts training times to 10 hours. The project features GeneFace++, an enhanced variant with improved synchronization and video quality, and MimicTalk for talking style control. It incorporates a PyTorch-based deep3D reconstruction module and pitch-aware audio-to-motion conversion. Visit the project page for more details and video demonstrations.