Skip to content

Latest commit

 

History

History
33 lines (11 loc) · 1.57 KB

File metadata and controls

33 lines (11 loc) · 1.57 KB

From Third Person to First Person: Dataset and Baselines for Synthesis and Retrieval

https://arxiv.org/abs/1812.00104

Accepted for presentation at EPIC@CVPR2019 workshop

ThirdToFirst

The entire real dataset can be found here: Real Dataset

The real dataset organized as the training/validation/testing splits: Real Splits

Their annotations can be found here: Real Data Annotations

The synthetic dataset organized as the training/validation/testing splits: Synthetic Splits

The generator used to syntheize data with the running instructions can be found here: Synthetic Dataset Generator

You can find the instructions for the synthetic generator in "Readme.txt".

Final Results can be found here: Final Results