Enhancing Video-Language Representations with Structural Spatio-Temporal Alignment

计算机科学 人工智能 自然语言处理 计算机视觉 模式识别(心理学)
作者
Hao Fei,Shengqiong Wu,Meishan Zhang,Min Zhang,Tat‐Seng Chua,Shuicheng Yan
出处
期刊:IEEE Transactions on Pattern Analysis and Machine Intelligence [IEEE Computer Society]
卷期号:46 (12): 7701-7719 被引量:14
标识
DOI:10.1109/tpami.2024.3393452
摘要

While pre-training large-scale video-language models (VLMs) has shown remarkable potential for various downstream video-language tasks, existing VLMs can still suffer from certain commonly seen limitations, e.g., coarse-grained cross-modal aligning , under-modeling of temporal dynamics, detached video-language view. In this work, we target enhancing VLMs with a fine-grained structural spatio-temporal alignment learning method (namely Finsta). First of all, we represent the input texts and videos with fine-grained scene graph (SG) structures, both of which are further unified into a holistic SG (HSG) for bridging two modalities. Then, an SG-based framework is built, where the textual SG (TSG) is encoded with a graph Transformer, while the video dynamic SG (DSG) and the HSG are modeled with a novel recurrent graph Transformer for spatial and temporal feature propagation. A spatial-temporal Gaussian differential graph Transformer is further devised to strengthen the sense of the changes in objects across spatial and temporal dimensions. Next, based on the fine-grained structural features of TSG and DSG, we perform object-centered spatial alignment and predicate-centered temporal alignment respectively, enhancing the video-language grounding in both the spatiality and temporality. We design our method as a plug&play system, which can be integrated into existing well-trained VLMs for further representation augmentation, without training from scratch or relying on SG annotations in downstream applications. On 6 representative VL modeling tasks over 12 datasets in both standard and long-form video scenarios, Finsta consistently improves the existing 13 strong-performing VLMs persistently, and refreshes the current state-of-the-art end task performance significantly in both the fine-tuning and zero-shot settings.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
2秒前
2秒前
JIE完成签到,获得积分20
2秒前
万能图书馆应助苏星星采纳,获得10
4秒前
情怀应助dd采纳,获得10
5秒前
Henry发布了新的文献求助10
6秒前
丘比特应助科研通管家采纳,获得10
6秒前
芝麻应助科研通管家采纳,获得10
6秒前
科研通AI5应助科研通管家采纳,获得30
6秒前
脑洞疼应助科研通管家采纳,获得10
6秒前
科研通AI5应助科研通管家采纳,获得10
6秒前
6秒前
wanci应助科研通管家采纳,获得10
6秒前
6秒前
6秒前
6秒前
6秒前
7秒前
7秒前
8秒前
9秒前
10秒前
10秒前
11秒前
bkagyin应助Omni采纳,获得10
11秒前
11秒前
夨艺发布了新的文献求助10
11秒前
jake完成签到,获得积分10
12秒前
水知寒完成签到,获得积分10
13秒前
13秒前
巴顿将军完成签到 ,获得积分10
13秒前
鹰击长空完成签到,获得积分10
14秒前
大个应助easy采纳,获得100
15秒前
16秒前
休123发布了新的文献求助10
16秒前
lixuerui发布了新的文献求助10
16秒前
苏星星发布了新的文献求助10
16秒前
邓佳鑫Alan应助EddyLalala采纳,获得10
17秒前
墚玊发布了新的文献求助10
18秒前
19秒前
高分求助中
Continuum Thermodynamics and Material Modelling 2000
Neuromuscular and Electrodiagnostic Medicine Board Review 1000
こんなに痛いのにどうして「なんでもない」と医者にいわれてしまうのでしょうか 510
いちばんやさしい生化学 500
The First Nuclear Era: The Life and Times of a Technological Fixer 500
岡本唐貴自伝的回想画集 500
A mandible of Pliosaurus brachyspondylus (Reptilia, Sauropterygia) from the Kimmeridgian of the Boulonnais (France) 300
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 物理 生物化学 纳米技术 计算机科学 化学工程 内科学 复合材料 物理化学 电极 遗传学 量子力学 基因 冶金 催化作用
热门帖子
关注 科研通微信公众号,转发送积分 3680654
求助须知:如何正确求助?哪些是违规求助? 3232950
关于积分的说明 9805307
捐赠科研通 2944164
什么是DOI,文献DOI怎么找? 1614528
邀请新用户注册赠送积分活动 762199
科研通“疑难数据库(出版商)”最低求助积分说明 737288