17 datasets found

A dataset of 3D objects, consisting of 36K image pairs with diverse backgrounds, lighting conditions and camera models.

Rel3D is a large-scale dataset of human-annotated spatial relations in 3D. It consists of spatial relations situated in synthetic 3D scenes, making it possible to extract rich...

3D object representation dataset

Relational CLEVR is a synthetic dataset of rendered 3D objects of various colours, shapes, sizes and textures.

Positional CLEVR is a synthetic dataset of rendered 3D objects of various colours, shapes, sizes and textures.

The YCB-V dataset is used for testing the class-adaptive object detector. It contains 33 objects with a large number of occurrences.

The dataset used for evaluating latent-space interpolation algorithms for variational autoencoders.

MPI3D is a benchmark dataset that consists of 1,036,800 examples of 3D objects from three different domains: Toy, RealisticC, and Real.

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.

3D-Chairs dataset consists of 30,000 training and 10,000 testing samples.

The NYU Depth V2 dataset contains two sub-datasets: labeled dataset and raw dataset. The raw dataset consists of 249 training scenes and 215 test scenes captured with Microsoft...

The MVTec ITODD dataset is a dataset for 3D object recognition in industry.

Zero-shot learning (ZSL) aims to classify objects that are not observed or seen during training. It relies on class semantic description to transfer knowledge from the seen...

3D Convolutional Neural Networks are sensitive to transformations applied to their input. This is a problem because a voxelized version of a 3D object, and its rotated clone,...

The Make3D dataset is a small but challenging dataset for 3D object recognition and scene understanding, which consists of 134 images with aligned depth information.

The ShapeNetCore dataset is a large-scale 3D model dataset, containing 44,000 3D models and 13 categories.

Point cloud registration is a crucial problem in computer vision and robotics. Existing methods either rely on matching local geometric features, which are sensitive to the pose...