Revitalize Region Feature for Democratizing Video-Language Pre-training of Retrieval

Abstract

Recent dominant methods for video-language pre-training (VLP) learn transferable representations from the raw pixels in an end-to-end manner to achieve advanced performance on downstream video-language tasks. Despite the impressive results, VLP research becomes extremely expensive with the need for massive data and a long training time, preventing further explorations. In this work, we revitalize region features of sparsely sampled video clips to significantly reduce both spatial and temporal visual redundancy towards democratizing VLP research at the same time achieving state-of-the-art results. Specifically, to fully explore the potential of region features, we introduce a novel bidirectional region-word alignment regularization that properly optimizes the fine-grained relations between regions and certain words in sentences, eliminating the domain/modality disconnections between pre-extracted region features and text. Extensive results of downstream text-to-video retrieval and video question answering tasks on seven datasets demonstrate the superiority of our method on both effectiveness and efficiency, e.g., our method achieves competing results with 80% fewer data and 85% less pre-training time compared to the most efficient VLP method so far [29]. The code will be available at https://github.com/showlab/DemoVLP.

Managing Metaverse Data Tsunami: Actionable Insights

“…Different metaverse users may have different tasks that require AI to assist, making it challenging to customize AI model and training data for each use case. A promising solution is to develop foundation deep learning models [96], [176]- [178] that are first trained on large-scale data and then can be applied to various downstream tasks with minimal task-specific fine-tuning. As a starting point, the recent work EgoVLP [179] conducted visual-language pretraining on Ego4D [180], a massive egocentric video dataset that targets collecting what smart glasses see throughout human daily activities, and then the pretrained EgoVLP model demonstrates strong performances on various metaverse applications such as action recognition, moment query via a textual description, object state change detection.…”

Learning Semantic Alignment with Global Modality Reconstruction for Video-Language Pre-training towards Retrieval

“…VideoBERT (Sun et al 2019b), ActBert (Zhu and Yang 2020), DECEMBERT (Tang, Lei, and Bansal 2021), and VIOLET (Fu et al 2021) pre-train matching tasks using the special token [CLS] for binary classification (Ruan and Jin 2022) with a cross-modal encoder (Vaswani et al 2017). Some methods (Zellers et al 2021;Ge et al 2022;Miech et al 2019Miech et al , 2020Ging et al 2020;Wang et al 2022b;Yang, Bisk, and Gao 2021;Yan et al 2021;Luo et al 2021;Patrick et al 2020;Cai et al 2022;Li et al 2020;Xu et al 2021b;Cai et al 2022;Cao et al 2022) pre-train matching tasks with two-stream encoders by forcing the paired samples closer while pushing different ones away (Ruan and Jin 2022). The others (Luo et al 2020;Li et al 2022) combine cross-modal Transformer matching tasks and twostream encoders matching tasks for more vital learning ability.…”