LSMDC

The LSMDC (Large Scale Movie Description Challenge) dataset contains video clips from movies paired with descriptive text, and it is used to evaluate text-video retrieval models by assessing their ability to match relevant text with corresponding video segments.

Papers using LSMDC (6)

Bima: Towards Biases Mitigation For Text-video Retrieval Via Scene Element Guidance2025 · 1 citesGAIS: Frame-level Gated Audio-visual Integration With Semantic Variance-scaled Perturbation For Text-video Retrieval2025CLIP-ViP: Adapting Pre-trained Image-Text Model to Video-Language Representation Alignment2022 · 53 citesX-Pool: Cross-Modal Language-Video Attention for Text-Video Retrieval2022 · 11 citesTagging before Alignment: Integrating Multi-Modal Tags for Video-Text Retrieval2023Mug-STAN: Adapting Image-Language Pretrained Models for General Video Understanding2023