LoVoRA: Text-guided and Mask-free Video Object Removal and Addition with Learnable Object-aware Localization Paper • 2512.02933 • Published Dec 2, 2025 • 4