《Unsupervised Learning of Depth and Ego-Motion from Video》读书笔记
2017-12-07 11:09
561 查看
原文：Unsupervised Learning of Depth and Ego-Motion from Video
实质：用单张图片推理场景结构：SfMLearner
相关作用：
针对端对端视觉里程计 ： 今年CVPR的 SfM-Learner。文章的核心思想是利用photometric consistency原理来估计每一帧的depth和pose。photometric consistency就是对于同一个物体的点，在不同两帧图像上投影点，图像灰度应该是一样的。各位回忆一下直接法SLAM的经典：LSD-SLAM (LSD-SLAM: Large-Scale Direct Monocular SLAM)，有没有感觉这篇文章的核心思路和LSD-SLAM如出一辙？本质都是优化photometric error。 来看看SfM-Learner的Loss（最终的Loss在此基础上做了优化），出自知乎。
他山之石：1. 用单张图片推理场景结构：UC Berkeley提出3D景深联合学习方法
2. UC伯克利联合谷歌推出无监督深度学习框架，模仿人眼实现视频中的自我运动认知
3. 专栏 | CVPR 2017 论文解读：基于视频的无监督深度和车辆运动估计
题目：基于视频的无监督深度和车辆运动估计
实质：用单张图片推理场景结构：SfMLearner
相关作用：
针对端对端视觉里程计 ： 今年CVPR的 SfM-Learner。文章的核心思想是利用photometric consistency原理来估计每一帧的depth和pose。photometric consistency就是对于同一个物体的点，在不同两帧图像上投影点，图像灰度应该是一样的。各位回忆一下直接法SLAM的经典：LSD-SLAM (LSD-SLAM: Large-Scale Direct Monocular SLAM)，有没有感觉这篇文章的核心思路和LSD-SLAM如出一辙？本质都是优化photometric error。 来看看SfM-Learner的Loss（最终的Loss在此基础上做了优化），出自知乎。
他山之石：1. 用单张图片推理场景结构：UC Berkeley提出3D景深联合学习方法
2. UC伯克利联合谷歌推出无监督深度学习框架，模仿人眼实现视频中的自我运动认知
3. 专栏 | CVPR 2017 论文解读：基于视频的无监督深度和车辆运动估计
题目：基于视频的无监督深度和车辆运动估计
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- 【ML】ICML2015_Unsupervised Learning of Video Representations using LSTMs
- Deep Learning 读书笔记（十五）：Deep Learning from Temporal Coherence in Video
- 每日论文 Learning from Simulated and Unsupervised Images through Adversarial Training
- Note for video Machine Learning and Data Mining——Theory of Generalization
- [Stereo_unsupervised][cvpr16]Unsupervised learning of disparity maps from stereo images
- What are the advantages of semi-supervised learning over supervised and unsupervised learning?
- 《Realtime and Robust Hand Tracking from Depth》读书笔记2
- Unsupervised Learning and Text Mining of Emotion Terms Using R
- More 3D Graphics (rgl) for Classification with Local Logistic Regression and Kernel Density Estimates (from The Elements of Statistical Learning)（转）
- 【转帖】UFLDL Tutorial（the main ideas of Unsupervised Feature Learning and Deep Learning）
- Some 3D Graphics (rgl) for Classification with Splines and Logistic Regression (from The Elements of Statistical Learning)（转）
- PredNet --- Deep Predictive coding networks for video prediction and unsupervised learning --- 论文笔记
- Postprocessing Depth Of Field and Real Motion Blur
- [Paper note] Learning from Simulated and Unsupervised Images through Adversarial Training
- Effective Java 英文 第二版 读书笔记 Item 13:Minimize the accessibility of classes and members
- 【JavaLearning】接口与抽象类的比较（The comparison of Interface and Abstract Class ）
- 读书笔记：Mastering the game of Go with deep neural networks and tree search
- ON THE EVOLUTION OF MACHINE LEARNING: FROM LINEAR MODELS TO NEURAL NETWORKS
- Machine Learning with Scikit-Learn and Tensorflow 7.3 Out-of-Bag评价方式
- A Taste of AOP from Solving Problems with OOP and Design Patterns (Part I)