Real-Weather World

Citation Author(s):
Hanyu
Zhou
Submitted by:
Hanyu Zhou
Last updated:
Thu, 08/22/2024 - 09:53
DOI:
10.21227/yaaj-0f05
Data Format:
*.avi; *.png; *.csv; *.zip
Links:
Unsupervised Cumulative Domain Adaptation for Foggy Scene Optical Flow
License:
Creative Commons Attribution
359 Views
Categories:
Artificial Intelligence
Machine Learning
Keywords:
optical flow
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

This is an optical flow dataset which covers multiple independent moving object samples under various adverse weather. This is an optical flow dataset which covers multiple independent moving object samples under various adverse weather. This is an optical flow dataset which covers multiple independent moving object samples under various adverse weather.

Instructions: 

For the collection, we obtain the video sequences from Youtube and camera, and then manually split them into paired adjacent images according to weather types and scene diversity. For the ground truth, with the help of human-assisted motion annotation method, we first segment the mask of the moving object, and then manually label keypoints in adjacent images for the moving object, thus generating the corresponding motion displacement as the ground truth.

Dataset Files

    Files have not been uploaded for this dataset

     

    QUESTIONS?