1,082 datasets found

MERCURY accelerates DNN training by exploiting input similarity.

A novel active solubility sensing device using computer vision is proposed to improve separation purification performance and prevent malfunctions of separation equipment.

The dataset used to test the proposed variable pixel imaging method.

The proposed parallel variable pixel arrays system for generating images gives an improved visual quality for the images studied.

This paper presents a benchmark for visual odometry and SLAM.

1 Year, 1000km: The Oxford RobotCar Dataset

The dataset used in this paper is a large dataset of simulated physical experiments, consisting of images of rolling balls on bowls of varying shape and orientation, and on...

Multi-view object image dataset (MVOD 5K)

Benchmark datasets for Generalized Zero-Shot Learning (GZSL) including AWA2, APY, CUB, and SUN.

The dataset is used to test the proposed method for building a relevance map through an autonomous exploration driven by a robotic arm.

The dataset is used for estimating the time-varying relative pose between two camera-IMU rigs on a flexible wing UAV.

A dataset of procedurally generated satellite images and their corresponding DEMs.

The dataset used in the paper for reflection symmetry detection in real 3D scans.

The dataset used in the paper for reflection symmetry detection in real images.

The dataset used in the paper for reflection symmetry detection in 2D and 3D point clouds.

A video data source, such as multispectral video acquired during administration of fluorescent tracers

A synthetic dataset for 3D lane detection, simulating abundant visual elements and focusing on evaluating a method's generalization capability to rarely observed scenarios.

The Taskonomy dataset contains 16 types of annotation, including object classification, scene classification, depth estimation, surface normals, curvature estimation, edge...

Endo-depth-and-motion: Reconstruction and tracking in endoscopic videos using depth networks and photometric constraints

Unsupervised monocular depth estimation models make use of adjacent frames as a supervisory signal during the training phase. However, temporally correlated frames are also...