CVPR 2018 值得关注的论文

本文探讨了深度学习领域的多项先进技术,包括图匹配的深度学习、基于排名的损失函数优化、深层聚合、关系网络用于目标检测、深度对抗度量学习等。文章还涉及了视觉问答与图像描述、视觉问题生成、图像与句子匹配的概念学习、Squeeze-and-Excitation网络、神经婴儿语等创新研究。此外,还讨论了图像生成、场景几何与语义的多任务学习、情感注意力、弱监督耦合网络、距离度量学习、上下文嵌入网络等。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

  • Deep Learning of Graph Matching
  • Efficient Optimization for Rank-Based Loss Functions
  • Deep Layer Aggregation
  • Relation Networks for Object Detection
  • Deep Adversarial Metric Learning
  • Link and Code: Fast Indexing With Graphs and Compact Regression Codes
  • Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
  • Visual Question Generation as Dual Task of Visual Question Answering
  • Learning Semantic Concepts and Order for Image and Sentence Matching
  • Squeeze-and-Excitation Networks
  • Neural Baby Talk
  • Revisiting Dilated Convolution: A Simple Approach for Weakly- and Semi-Supervised Semantic Segmentation
  • Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics
  • Emotional Attention: A Study of Image Sentiment and Visual Attention
  • Weakly Supervised Coupled Networks for Visual Sentiment Analysis
  • Large-Scale Distance Metric Learning With Uncertainty
  • Context Embedding Networks
  • Iterative Learning With Open-Set Noisy Labels
  • Interleaved Structured Sparse Convolutional Neural Networks
  • VITAL: VIsual Tracking via Adversarial Learning
  • Detail-Preserving Pooling in Deep Networks
  • Deep Mutual Learning
  • Image Generation From Scene Graphs
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值