人工智能培训

搜索

机器学习论文:视觉语义信息追求:一项调查(Visual Semantic Information Pursuit: A Survey)

[复制链接]
mhtq 发表于 2019-3-15 13:02:18 | 显示全部楼层 |阅读模式
mhtq 2019-3-15 13:02:18 217 0 显示全部楼层
机器学习论文:视觉语义信息追求:一项调查(Visual Semantic Information Pursuit: A Survey)视觉语义信息包括两个重要部分:每个视觉语义单元的含义和由这些视觉语义单元传达的连贯的视觉语义关系。本质上,前者是视觉感知任务,而后者则是视觉上下文推理。由于深度学习的成功,已经实现了视觉感知的显着进步。相比之下,视觉语义信息追求,一种结合视觉感知和视觉上下文推理的视觉场景语义解释任务,仍处于早期阶段。它是许多不同计算机视觉应用的核心任务,例如对象检测,视觉语义分割,视觉关系检测或场景图生成。由于它有助于提高所得到的解释的准确性和一致性,因此视觉上下文推理通常与当前深度端到端视觉语义信息追踪方法中的视觉感知相结合。然而,对这个令人兴奋的领域进行全面审查仍然是一个问题。在本次调查中,我们为所有这些方法提供了统一的理论范例,然后概述了每个潜在方向的主要发展和未来趋势。还介绍了常用的基准数据集,评估指标和相应方法的比较。
Visual semantic information comprises two important parts: the meaning ofeach visual semantic unit and the coherent visual semantic relation conveyed bythese visual semantic units.Essentially, the former one is a visual perceptiontask while the latter one corresponds to visual context reasoning.Remarkableadvances in visual perception have been achieved due to the success of deeplearning.In contrast, visual semantic information pursuit, a visual scenesemantic interpretation task combining visual perception and visual contextreasoning, is still in its early stage.It is the core task of many differentcomputer vision applications, such as object detection, visual semanticsegmentation, visual relationship detection or scene graph generation.Since ithelps to enhance the accuracy and the consistency of the resultinginterpretation, visual context reasoning is often incorporated with visualperception in current deep end-to-end visual semantic information pursuitmethods.However, a comprehensive review for this exciting area is stilllacking.In this survey, we present a unified theoretical paradigm for allthese methods, followed by an overview of the major developments and the futuretrends in each potential direction.The common benchmark datasets, theevaluation metrics and the comparisons of the corresponding methods are alsointroduced.机器学习论文:视觉语义信息追求:一项调查(Visual Semantic Information Pursuit: A Survey) CJd6M8tJzboO9j89.jpg
URL地址:https://arxiv.org/abs/1903.05434     ----pdf下载地址:https://arxiv.org/pdf/1903.05434    ----机器学习论文:视觉语义信息追求:一项调查(Visual Semantic Information Pursuit: A Survey)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

mhtq当前离线
新手上路

查看:217 | 回复:0

快速回复 返回顶部 返回列表