ML Research Hub
32.8K subscribers
4.13K photos
244 videos
23 files
4.46K links
Advancing research in Machine Learning – practical insights, tools, and techniques for researchers.

Admin: @HusseinSheikho || @Hussein_Sheikho
Download Telegram
Media is too big
VIEW IN TELEGRAM
Generative Video Motion Editing with 3D Point Tracks

📝 Summary:
This paper presents a track-conditioned video-to-video framework for precise joint camera and object motion editing. It uses 3D point tracks to maintain spatiotemporal coherence and handle occlusions through explicit depth cues. This enables diverse motion edits.

🔹 Publication Date: Published on Dec 1

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2512.02015
• PDF: https://arxiv.org/pdf/2512.02015
• Project Page: https://edit-by-track.github.io/

==================================

For more data science resources:
https://t.me/DataScienceT

#VideoEditing #GenerativeAI #ComputerVision #3DTracking #DeepLearning
1👍1