The paper introduces a massive dataset of gait (walking patterns) extracted from over 1,000 YouTube videos to help train machine learning models for person identification.
The naming convention patel_7_7-1 typically refers to the 7th subject and the 7th video sequence within the dataset's internal organization.
The primary academic paper that covers this video and the associated research is: patel_7_7-1mp4
(Search for "YouTube-Gait Patel") ResearchGate
You can find the full text and technical specifications for the YouTube-Gait dataset through these research repositories: The paper introduces a massive dataset of gait
(Pre-print versions are often available here for free)
by Ajay Kumar Patel et al. Key Details from the Research Key Details from the Research The authors used
The authors used Gait Energy Images (GEI) —a temporal average of binary silhouettes of a walking person—to represent the unique walking style of each individual in the video. Where to Access the Paper