68 datasets found

A real-world dataset captured by mobile phones and containing quite complex trajectories.

A novel end-to-end method that does not require prior camera parameters to deal with image sequences with complex trajectories.

The Kubric dataset offers training and validation data across various difficulty levels, featuring videos of objects interacting with each other.

The dataset used for training and testing the proposed RGBD-based obstacle avoidance system for visually impaired people.

The KITTI Vision Benchmark Suite is a dataset used for object detection and tracking in autonomous vehicles.

Zero-1-to-3: Zero-shot one image to 3D object.

The Middlebury dataset is a benchmark for stereo vision and 3D reconstruction.

Shapenet is a large-scale synthesis 3D object dataset, where we follow [9] to use the official test splits of chair, car, and motorbike categories for evaluation since they...

DeepFashion3D is a dataset for 3D garment reconstruction from single images.

Multi-task learning (MTL) research is broadly divided into two categories: one is to learn the correlation between tasks through model structures, and the other is to balance...

The Pix3D dataset is a collection of 3D models with RGB images, used for training and testing 3D object recognition and generation tasks.

Neural Radiance Fields (NeRFs) model a 3D scene as a volumetric function, which can be rendered from arbitrary viewpoints to generate highly-realistic images.

The DTU MVS Dataset and the Local Light Field Fusion Dataset are used to evaluate the performance of the proposed GARF model.

The ScanNet dataset is a large-scale indoor dataset composed of monocular sequences with ground truth poses and depth images.

Semantic segmentation is a crucial and challenging task for image understanding. It aims to predict a dense labeling map for the input image, which assigns each pixel a unique...

The Objaverse dataset contains around 800k 3D objects. After adopting simple filter leveraging CLIP [27] to remove the objects whose rendered images are not relevant to its...

The dataset used in the paper SpikingNeRF: Making Bio-inspired Neural Networks See through the Real World

The dataset used in the paper SpikingNeRF: Making Bio-inspired Neural Networks See through the Real World

The dataset FAUST contains 10 different bodies in 10 different poses. The irregular surface meshes have 6890 vertices. The dataset is used to evaluate the reconstruction quality...

NeRF [33] has demonstrated amazing ability to synthesize images of 3D scenes from novel views. However, they rely upon specialized volumetric rendering algorithms based on ray...