Tohoku-University-6D-Pose-Estimation-Dataset
Pose EstimationIndustrial Automation
This is an open dataset for evaluating pose‑estimation algorithms on industrial parts. The scenes simulate real factory conditions, containing a single type of part randomly stacked, with more than ten parts per scene, typically made of resin, lacking distinct colour or texture, and having similar shapes. The dataset is divided into simulated and real‑world scenes; currently only simulated data have been uploaded, with more data under preparation.
Source githubUpdated Apr 24, 2020173 viewsLinked
Inspect dataset