29 datasets found

The dataset used for monocular depth estimation task

The dataset used for monocular depth estimation task

The dataset used for training and testing the model consists of 3D walking motion capture data of three different participants. The data is edited to form a continuous cycle...

The SUNCG dataset is a large dataset of 3D indoor scenes with missing objects. The dataset is used to evaluate the performance of scene augmentation and context-based object...

The Learning to predict 3D surfaces of sculptures from single and multiple views dataset is a dataset for predicting 3D shapes of sculptures from a single or multiple images.

The Through the Looking Glass dataset is a dataset for predicting 3D shapes of transparent shapes from a single image.

SinGAN is a new unconditional generative model that can be learned from a single natural image.

The dataset used for training and testing the proposed TetraTSDF model for 3D human reconstruction from a single image with a tetrahedral outer shell.

Depth completion from a sparse depth map guided by a single image. Our method generates a smooth depth map while preserving discontinuity between different objects.

The LRS2 dataset consists of 48,164 video clips from outdoor shows on BBC television. Each video is accompanied by an audio corresponding to a sentence with up to 100 characters.

Camouflaged object detection (COD) aims to search and segment camouflaged objects from single image.

The CAMO dataset is the first COD dataset, containing 1,250 camouflaged images and 1,250 non-camouflaged images.

The NYUv2 dataset is a large-scale dataset for 3D object recognition and semantic segmentation. It contains 206 test set video sequences with 135 classes.

The dataset is created by augmenting a single input image with masks that indicate the presence of target concepts. The masks can be provided by the user or generated...

The WorldExpo’10 dataset contains 199,923 pedestrians labeled at their centers of heads.

The Break-A-Scene dataset contains images with multiple concepts extracted from a single image.

The COD10K dataset is currently the largest challenging dataset for COD, containing 10K images with dense annotations.

The dataset used for training and testing the direct regression model, sparse correspondence model, and dense correspondence model.

The dataset used for learning-based methods for online eye-in-hand camera calibration from a single RGB image.

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