WMRemove

Hey everyone! 👋 I’m Mohammed, the creator of WMRemove. The Inspiration: The explosion of generative video tools like Sora and Runway ML has been incredible to watch, but I noticed a recurring friction point: amazing clips were often unusable for professional mockups or client storyboards because of the pervasive watermarks. I wanted to build a bridge that allows creators to use these tools freely in their production workflows without the visual distraction. The Problem: Most existing "removers" just blindly blur a rectangle. That works for a static logo on a black screen, but it fails miserably on complex, moving shots. I wanted to solve the "Dynamic Watermark" problem—where a logo drifts across a detailed background—without requiring the user to edit frame-by-frame. The Evolution: We actually started with a simple FFmpeg wrapper, but quickly realized it wasn't enough. The project evolved significantly as we integrated Computer Vision (Optical Flow) to track moving watermarks automatically. We then layered on AI Inpainting to reconstruct the pixels rather than just hiding them. It’s been a journey from a simple command-line script to a full-stack editing suite! I’d love for you to try the Auto Mode and let me know how it handles your clips. We are constantly tuning the tracking algorithms, so your feedback is gold. 🚀 Thanks for checking it out!