VideoGLaMM: A Large Multimodal Model for Pixel-Level Visual Grounding in Videos

Shehan Munasinghe, Hanan Gani, Wenqi Zhu +4

Fine-grained alignment between videos and text is challenging due to complex spatial and temporal dynamics in videos. Existing video-based Large Multimodal Models (LMMs) handle bas…