15 datasets found

Implicit neural 3D representation has achieved impressive results in surface or scene reconstruction and novel view synthesis, which typically uses the coordinate-based...

Consistent-1-to-3 is a novel framework that generates consistent images of any objects from any viewpoint given a single image.

A large-scale outdoor multi-modal dataset and benchmark for novel view synthesis and implicit scene reconstruction.

NeRF dataset for large-scale urban scene reconstruction

The LLFF-NeRF dataset consists of eight scenes captured with a handheld cellphone, with 20-62 images each. The resolution of each image in the dataset is 4032 × 3024.

The dataset used in the paper for 3D geometry reconstruction and novel view synthesis.

The dataset used in the paper for 3D geometry reconstruction and novel view synthesis.

A system for 3D reconstruction and novel view synthesis from an arbitrary number of unposed images.

The dataset used in the paper is the NVIDIA Dynamic Scene Dataset, which contains 8 diverse scenes with 12 sequences captured by synchronized cameras at fixed positions.

The dataset used for training and testing the Uncertainty-guided Optimal Transport (UGOT) approach for depth supervision in sparse-view 3D Gaussian splatting for novel view...

The dataset used in the paper is the LLFF dataset, which contains real-world scenes and is used for training and testing the proposed neural radiance field model.

The DTU, BlendMVS, and H3DS datasets are used for evaluating the performance of the VQ-NeRF method.

A method that synthesizes novel views of complex scenes by interpolating a sparse set of nearby views.

The Mip-NeRF360 dataset contains scenes taken from a 360 degree view, with emphasis on minimizing photometric variations.

This paper introduces a new co-regularization perspective for improving sparse-view 3DGS. We observe the two 3D Gaussian radiance fields exhibit different behaviors for the same...