Explore a 14-minute conference presentation from SIGIR 2024 that introduces M2-RAAP, an innovative multi-modal approach for advancing adaptation-based pre-training in zero-shot video-text retrieval. Learn about the methodology developed by researchers Xingning Dong, Zipeng Feng, Chunluan Zhou, Xuzheng Yu, Ming Yang, and Qingpei Guo that aims to enhance the effectiveness and efficiency of matching video content with textual descriptions without prior training examples. Understand how this cutting-edge research presented at the Association for Computing Machinery (ACM) conference contributes to the field of multimedia information retrieval and machine learning.
M2-RAAP: A Multi-Modal Recipe for Advancing Adaptation-based Pre-training for Video-text Retrieval
Association for Computing Machinery (ACM) via YouTube
Overview
Syllabus
SIGIR 2024 W1.6 [fp] M2-RAAP: A Multi-Modal Recipe for Advancing Adaptation-based Pre-training
Taught by
Association for Computing Machinery (ACM)
Related Courses
-
Fine-grained Textual Inversion Network for Zero-Shot Composed Image Retrieval - Multimedia 1
-
Uncertainty-Guided Noisy Correspondence Learning for Efficient Cross-Modal Matching - Multimodal Session 3.3
-
Drop Your Decoder: Pre-training with Bag-of-Word Prediction for Dense Passage Retrieval
-
CFIR: Fast and Effective Long-Text to Image Retrieval for Large Corpora
-
Generative Retrieval as Multi-Vector Dense Retrieval
-
IM-RAG: Multi-Round Retrieval-Augmented Generation Through Learning Inner Monologues
