See More, Store Less: Memory-Efficient Resolution for Video Moment Retrieval
Authors:
- Minchol KwonMKMinchol Kwon
- Jonghee KimJKJonghee Kim
- Junyeong KimJKJunyeong Kim
Preprint
Abstract
Recent advances in Multimodal Large Language Models (MLLMs) have improved image recognition and reasoning, but video-related tasks remain challenging due to memory constraints from dense frame processing. Existing Video Moment Retrieval (VMR) methodologies rely on sparse frame sampling, risking potential information loss, especially in lengthy videos. We propose SMORE (See MORE, store less), a framework that enhances memory efficiency while maintaining high information resolution. SMORE (1) uses query-guided captions to encode semantics aligned with user intent, (2) applies query-aware importance modulation to highlight relevant segments, and (3) adaptively compresses frames to preserve key content while reducing redundancy. This enables efficient video understanding without exceeding memory budgets. Experimental validation reveals that SMORE achieves state-of-the-art performance on QVHighlights, Charades-STA, and ActivityNet-Captions benchmarks.Keywords
All Related Versions
This publication has 6 references indexed in Scilit:
- Number it: Temporal Grounding Videos like Flipping Manga2025
- VisionZip: Longer is Better but Not Necessary in Vision Language Models2025
- InternVideo2: Scaling Foundation Models for Multimodal Video Understanding2024
- UniMD: Towards Unifying Moment Retrieval and Temporal Action Detection2024
- InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning2023
- CLIPScore: A Reference-free Evaluation Metric for Image Captioning2021