人物交互(human object interaction)论文汇总-2020年
1. Learning Human-Object Interaction Detection using Interaction Points 1.1 总述 大多数现有的HOI检测方法都是以实例为中心的,其中基于外
HOTR: End-to-End Human-Object Interaction Detection with Transformers论文阅读笔记
一、本文的内容 1. 研究目的 本文提出了一种基于transformer的人物交互的新的框架,它能够根据图像预测出a pair of 三元组( 人,物,交互)&#x
论文《AutoFIS: Automatic Feature Interaction Selection in Factorization Models for Click-Through R..》阅读
论文《AutoFIS: Automatic Feature Interaction Selection in Factorization Models for Click-Through Rate Prediction》阅读 AutoFIS
论文阅读-Modular Interactive Video Object Segmentation Interaction-to-Mask, Propagation
Abstract 我们提出了模块化交互式 VOS (MiVOS) 框架,该框架将interaction-to-mask和mask propagation解耦,从而实现更高的通用性和更好的性能。单
Cascaded Human-Object Interaction Recognition论文阅读笔记
笔记 现有的方法大都采用single-stage的推理线,考虑到任务的复杂性,作者提出了一种采用级联结构,多分支,从粗糙到细致的HOI理解。如图1,作者的模型包含了一个实例定位网络和一个交互识别网络。这两个网络都以级联的形式工作,通过实例定
Learning Attentive Pairwise Interaction for Fine-Grained Classification论文解读
论文链接:https:arxivabs2002.10191 分享的这篇文章来自于AAAI2020,文章的整个思路并不难理解。文章的idea来自于我们人类对相似图像的识别。一般来说,我们识别相似的图像,一方面是去找到图像中特殊的区域
Modular Interactive Video Object Segmentation: Interaction-to-Mask, Propagation and Difference-Aware
TitleAuthorsPubLinkModular Interactive Video Object Segmentation:Interaction-to-Mask, Propagation and Difference-Aware F
论文笔记 ACL 2021|Capturing Event Argument Interaction via A Bi-Directional Entity-Level Recurrent Decod
文章目录 1 简介1.1 动机1.2 创新 2 方法3 实验 1 简介 论文题目:Capturing Event Argument Interaction via A Bi-Directional Entity-Le
Channel Interaction Networks for Fine-Grained Image Categorization-笔记
沉睡在草稿箱的笔记 摘要 我们发现通道之间的细微差别可以帮助我们捕获细粒度识别所需要的细微差别。我们提出了一个 CIN channel interaction network,它可以捕获图像与图像之间的通道差距。对于
AN INTERACTION-AWARE ATTENTION NETWORK FOR SPEECH EMOTION RECOGNITION IN SPOKEN DIALOGS -情感识别论文学习
AN INTERACTION-AWARE ATTENTION NETWORK FOR SPEECH EMOTION RECOGNITION IN SPOKEN DIALOGS 简介构架结论 简介 构架 结论 针对该文章提出的结果&
Danmaku: A New Paradigm of Social Interaction via Online Videos作者的两篇论文核心概括
题目:Danmaku: A New Paradigm of Social Interaction via Online Videos Danmaku vs. Forum Comments: Understanding
【论文】(IJCAI20 知识图谱神经网络)KGNN: Knowledge Graph Neural Network for Drug-Drug Interaction Prediction
KGNN: Knowledge Graph Neural Network for Drug-Drug Interaction Prediction 背景相关研究主流方法【特点】:多数据源的集成+流行的嵌入方法【缺点】:对药物**与靶点和基因
论文解读:KGNN: Knowledge Graph Neural Network for Drug-Drug Interaction Prediction
一、背景 药物间的相互作用(DDI)是指同时或先后服用两种或两种以上药物时,药物之间所产生的相互作用,而该相互作用可能会导致意想不到的副作用。 总结归纳现有DDI预测方法,大
论文解读:SumGNN: Multi-typed Drug Interaction Prediction via Efficient Knowledge Graph Summarization(Bi)
Part1 Introduction 以前的大多数工作都集中在二元 DDI 预测上,而多类型 DDI 药理作用预测更有意义但任务更艰巨 SumGNN: knowledge summarization graph neu
论文笔记 ACL 2021|Document-level Event Extraction via Heterogeneous Graph-based Interaction Model with a
文章目录 1 简介1.1 动机1.2 创新 2 方法3 实验 1 简介 论文题目:Document-level Event Extraction via Heterogeneous Graph-based Inter
文献笔记|知识追踪|GIKT: A Graph-based Interaction Model for Knowledge Tracing
GIKT: A Graph-based Interaction Model for Knowledge Tracing 作者:Yang Yang, Jian Shen, Yanru Qu, Yunfei Liu, Ke
论文阅读:Compositional Learning for Human Object Interaction
Compositional Learning for HOI(ECCV 2018) 文章 作者的的想法是因为我们很难搜集到所有组合之间的interaction,所以必须会面临的问题就是要识别在数据集中从未见到过的情况,也就是HOI的z
Object Interaction Diagram -- OID
Definition: OIDs show interaction in your use case. More formally, how to organize the object interaction that take pla
IANet:Interaction-and-Aggregation Network for Person Re-identification阅读笔记
IANet:Interaction-and-Aggregation Network for Person Re-identification 1. 摘要 由于CNN具有固定的几何结构(卷积固定的滑动窗口),因此在模
RSIS 系列 Rotated Multi-Scale Interaction Network for Referring Remote Sensing Image Segmentation 论文阅读
RSIS 系列 Rotated Multi-Scale Interaction Network for Referring Remote Sensing Image Segmentation 论文阅读笔记 一、Abstract二、引言三、相
发表评论