Image-based localization using LSTMs for structured feature correlation
文中提出很多场景使用deeplearning的方式还没有传统的sift的效果好 本文主要是解决两种场景：1：无纹理。2:重复纹理 也是一个end-to-end的直接回归得到6dof pose的文章，只不过不仅仅使用了CNN。文章是CNN+LSTM的方式。 仍然使用pose 误差： 整体网路框架： 使用预训练的网络googlenet，同posenet一样最后的fc层输出的是2048维度的gl
相关文章
- 1. PredRNN: Recurrent Neural Networks for Predictive Learning using Spatiotemporal LSTMs
- 2. Correlation Congruence for Knowledge Distillation
- 3. NR-SIQA Using Convolutional Neural Network for Adaptive Feature Extraction
- 4. Linguistically Regularized LSTMs for Sentiment Classification
- 5. 【OverFeat】《OverFeat：Integrated Recognition, Localization and Detection using Convolutional Networks》
- 6. Visual Object Tracking using Adaptive Correlation Filters
- 7. R programming for feature selection and regression
- 8. OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
- 9. 【姿态估计文章阅读】Structured Feature Learning for Pose Estimation
- 10. 论文阅读：《Structured Feature Learning for Pose Estimation》CVPR 2016
- 更多相关文章...
- • Swift for 循环 - Swift 教程
- • Scala for循环 - Scala教程
- • RxJava操作符（六）Utility
- • C# 中 foreach 遍历的用法
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. PredRNN: Recurrent Neural Networks for Predictive Learning using Spatiotemporal LSTMs
- 2. Correlation Congruence for Knowledge Distillation
- 3. NR-SIQA Using Convolutional Neural Network for Adaptive Feature Extraction
- 4. Linguistically Regularized LSTMs for Sentiment Classification
- 5. 【OverFeat】《OverFeat：Integrated Recognition, Localization and Detection using Convolutional Networks》
- 6. Visual Object Tracking using Adaptive Correlation Filters
- 7. R programming for feature selection and regression
- 8. OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
- 9. 【姿态估计文章阅读】Structured Feature Learning for Pose Estimation
- 10. 论文阅读：《Structured Feature Learning for Pose Estimation》CVPR 2016