Get Started
Home
Topics
Search
Library
Research questionHow can partially relevant video retrieval locate precise query-relevant moments in untrimmed videos with weak supervision?Relevant moments may occupy only a small part of an untrimmed video, while weak supervision can make frame-level matches imprecise. The challenge is to distinguish and retrieve the moments that actually support the text query.
AI
Computer Vision
Image & Video Processing
Information Retrieval
Machine Learning
Multimodal Models
Latest papersRecent research connected to this question, newest first.Intrinsic Temporal Adaptation of CLIP for Partially Relevant Video RetrievalThe source studies partially relevant video retrieval using CLIP-based frame features and reports temporal adaptation and affinity-weighted learning for this setting. Evidence is provided through PRVR benchmarks, cross-dataset transfer, and frame-level retrieval accuracy within ground-truth relevant moments.research paper · Sep 4, 2026
Related questions
How can text-video retrieval preserve temporal structure when matching queries to heterogeneous video frames?How can text-promptable video segmentation track targets through disappearance while rejecting visually similar artifacts?How can we localize and describe multiple events in untrimmed videos from only ordered event captions?How can weakly supervised video anomaly detection adapt to anomalies with varying durations and temporal dynamics?