Explore a cutting-edge research presentation on Uncertainty-Guided Noisy Correspondence Learning (UGNCL) for efficient cross-modal matching. Delve into the innovative approach presented by authors Quanxing Zha, Xin Liu, Yiu-Ming Cheung, Xing Xu, Nannan Wang, and Jianjia Cao at the SIGIR 2024 conference. Learn about the challenges and solutions in multimodal matching, focusing on how UGNCL addresses noisy correspondences and improves efficiency in cross-modal applications. Gain insights into the latest advancements in information retrieval and machine learning techniques during this 11-minute talk organized by the Association for Computing Machinery (ACM).
Uncertainty-Guided Noisy Correspondence Learning for Efficient Cross-Modal Matching - Multimodal Session 3.3
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
SIGIR 2024 M3.3 [fp] UGNCL: Uncertainty-Guided Noisy Correspondence Learning for Efficient CM Match
Taught by
Association for Computing Machinery (ACM)
Related Courses
-
Unsupervised Cross-Domain Image Retrieval with Semantic-Attended Mixture-of-Experts
-
M2-RAAP: A Multi-Modal Recipe for Advancing Adaptation-based Pre-training for Video-text Retrieval
-
Self-Improving Teacher Cultivates Better Student: Distillation Calibration for Multimodal Large Language Models - Lecture 3.3
-
Steering Large Language Models for Cross-lingual Information Retrieval - Multilingual Retrieval
-
Multimodality Invariant Learning for Multimedia-Based New Item Recommendation - Lecture 6
-
Fine-grained Textual Inversion Network for Zero-Shot Composed Image Retrieval - Multimedia 1
