Meta AI Presents MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding

Meta AI Presents MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding