992 datasets found

The Mutant and LEGO dataset is a dynamic scene dataset. It contains 90% images for training and 10% images for evaluation.

The Tanks and Temples Advanced (T&T) dataset is a benchmark dataset for image-based 3D reconstruction. It contains 90% images for training and 10% images for evaluation.

The Visual Wake Words (VWW) dataset consists of high-resolution images that include visual cues to 'wake-up' AI-powered home assistant devices.

This paper introduces a horizontally scalable vision transformer (HSViT) scheme with a novel image-level feature embedding. The design of HSViT preserves the inductive bias from...

The dataset used in this paper is ADE20k for semantic segmentation.

The SPL2018 and DsTok datasets for computer-generated image detection

The proposed dual stream convolutional neural network for computer-generated image detection

The dataset used in this paper is a large variety of popular CNN models, such as straight-forward, complicated-connected, and grouped architectures.

The dataset used in the paper is ModelNet40 and ModelNet10, which are subsets of ShapeNet.

The dataset used in the paper is ShapeNet, a large-scale dataset of 3D models, and ModelNet40 and ModelNet10, which are subsets of ShapeNet.

Depth prediction is fundamental for many useful applications on computer vision and robotic systems. On mobile phones, the performance of some useful applications as augmented...

MobileNets, MnasNets, EfficientNets, and ResNets

Residual Networks (ResNet) is composed of stacked entities referred to as residual blocks. A Residual Block of ResNet contains a module and an identity loop.

The dataset used in this paper is a real-world vehicle point cloud collected from a real vehicle self-driving process.

Real-time intelligent detection and prediction of subjects' behavior particularly their movements or actions is critical in the ward.

A large-scale, physically unconstrained gaze estimation dataset collected from 238 subjects in indoor and outdoor settings.

Motion synthesis in real-world 3D scenes has recently attracted much attention. However, the static environment assumption made by most current methods usually cannot be...

DepthP+P: A method for self-supervised monocular depth estimation using planar and parallax.

The dataset used in this paper is a collection of 3D point clouds.

The NetVLAD dataset is used for evaluating the performance of visual representation learning methods.