今日CS.CV计算机视觉论文速览
Mon, 4 Feb 2019
Totally 27 papers
Interesting:
-
基于深度先验的超光谱图像去噪、修复和超分辨，利用CNN固有的性质，研究人员在深度先验的基础上克服了超光谱图像数据有限的问题，三维卷积可以将低层次信息有效的转换为先验，由于图像修复。(from 挪威科技大学)
二维和三维的架构如下图所示：
最后的一些效果
与相关方法的比较：
代码：https://github.com/acecreamu/deep-hs-prior
data:
Denoise: HYDICE DC Mall data[2]
Inpainting: Indian Pines dataset[3]
Super-resolution:ROSIS-03 image of Pavia Center [4] -
ColorNet, 探索了颜色空间对于分类任务的重要性，在模型中使用多种颜色空间可以在一定程度上提高分类准确率，达到同样的精度需要的参数也会减少。(from 清华大学)
不同颜色空间在分类任务下的混淆矩阵：
新提出的多种颜色空间混合模型：
-
遥感数据中的森林/非森林分类，（from 意大利DIETI, University Federico II）
一些结果：
数据集：tandem-x
“The global forest/non-forest map from tandem-x interferometric SAR data”
“High-resolution forest mapping from tandemx interferometric data exploiting nonlocal filtering” -
US-NET，基于同一个Unet基础网络，利用检测（定位分类）和分割（mask）网络共享输出融合互相提升效果，实现了对于病理切片细胞核较好的示例分割。(from Queen Mary University of London)
检测结果：
DataSet：
Segmentation of Nuclei in Images Contest??
MICCAI MoNuSeg -
梯度/拉普拉斯的逆过程，用于梯度域的图像编辑，提出了一种快速准确鲁棒的方法从梯度或拉普拉斯域逆计算出图像的方法，这一方法基于单卷积的格林函数并可以被GOU和快速傅里叶变换优化提速。文章还利用这一方法在泊松融合、梯度移除、保留边缘人物中进行了测试。（from 蒙特利尔理工）
一种用于图像重建的例子：
-
基于投影的2.5D U-Net架构，用于快速体分割。为了避免三维卷积庞大的计算量，使用投影将体数据投影为一系列图像序列，每幅图中包含了全部的数据信息；随后利用2D卷积来处理这些图像，并再次转回三维数据。(from University of Innsbruck)
网络架构如下：
从单个分割图像重建三维数据算子(maximum intensity projection,MIP) ：
Daily Computer Vision Papers
[1] *Title: Learning Differentiable Grammars for Continuous Data
Authors:AJ Piergiovanni, Anelia Angelova, Michael S. Ryoo
[2] *Title: Top-view Trajectories: A Pedestrian Dataset of Vehicle-Crowd Interaction from Controlled Experiments and Crowded Campus
Authors:Dongfang Yang, Linhui Li, Keith Redmill, Ümit Özgüner
[3] Title: Do we train on test data? Purging CIFAR of near-duplicates
Authors:Björn Barz, Joachim Denzler
[4] Title: Learnable Embedding Space for Efficient Neural Architecture Compression
Authors:Shengcao Cao, Xiaofang Wang, Kris M. Kitani
[5] Title: Self-Supervised Visual Representations for Cross-Modal Retrieval
Authors:Yash Patel, Lluis Gomez, Marçal Rusiñol, Dimosthenis Karatzas, C.V. Jawahar
[6] Title: The Generation and Application of Medical Image Grid Based on Differential Geometric Features
Authors:Yongpei Zhu, Zicong Zhou, Guojun Liao, Qianxi Yang, Kehong Yuan
[7] *Title: Projection-Based 2.5D U-net Architecture for Fast Volumetric Segmentation
Authors:Christoph Angermann, Markus Haltmeier, Ruth Steiger, Sergiy Pereverzyev Jr, Elke Gizewski
[8] Title: SensitiveNets: Learning Agnostic Representations with Application to Face Recognition
Authors:Aythami Morales, Julian Fierrez, Ruben Vera-Rodriguez
[9] Title: Rethinking Visual Relationships for High-level Image Understanding
Authors:Yuanzhi Liang, Yalong Bai, Wei Zhang, Xueming Qian, Li Zhu, Tao Mei
[10] * Title: Generative Smoke Removal
Authors:Oleksii Sidorov, Congcong Wang, Faouzi Alaya Cheikh
[11] *Title: Deep Hyperspectral Prior: Denoising, Inpainting, Super-Resolution
Authors:Oleksii Sidorov, Jon Yngve Hardeberg
[12] Title: End-to-end Lane Detection through Differentiable Least-Squares Fitting
Authors:Bert De Brabandere, Wouter Van Gansbeke, Davy Neven, Marc Proesmans, Luc Van Gool
[13] *Title: Deep Learning Solutions for TanDEM-X-based Forest Classification
Authors:Antonio Mazza, Francescopaolo Sica
[14] *Title: ColorNet: Investigating the importance of color spaces for image classification
Authors:Shreyank N Gowda, Chun Yuan
[15] Title: A Classification Supervised Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids
Authors:Qiuyu Zhu, Ruixin Zhang
[16] *Title: Fast and Optimal Laplacian Solver for Gradient-Domain Image Editing using Green Function Convolution
Authors:Dominique Beaini, Sofiane Achiche, Fabrice Nonez, Olivier Brochu Dufour, Cédric Leblond-Ménard, Mahdis Asaadi, Maxime Raison
[17] Title: Dataset Culling: Towards Efficient Training Of Distillation-Based Domain Specific Models
Authors:Kentaro Yoshioka, Edward Lee, Simon Wong, Mark Horowitz
[18] Title: Lift-the-Flap: Context Reasoning Using Object-Centered Graphs
Authors:Mengmi Zhang, Jiashi Feng, Karla Montejo, Joseph Kwon, Joo Hwee Lim, Gabriel Kreiman
[19] Title: Deep Triplet Quantization
Authors:Bin Liu, Yue Cao, Mingsheng Long, Jianmin Wang, Jingdong Wang
[20] *Title: US-net for robust and efficient nuclei instance segmentation
Authors:Zhaoyang Xu, Faranak Sobhani, Carlos Fernandez Moro, Qianni Zhang
[21] Title: Episodic Training for Domain Generalization
Authors:Da Li, Jianshu Zhang, Yongxin Yang, Cong Liu, Yi-Zhe Song, Timothy M. Hospedales
[22] *Title: Geometric Interpretation of side-sharing and point-sharing solutions in the P3P Problem
Authors:Bo wang, Hao Hu, Caixia Zhang
[23] *Title: Learning Metric Graphs for Neuron Segmentation In Electron Microscopy Images
Authors:Kyle Luther, H. Sebastian Seung
[24] Title: BLOCK: Bilinear Superdiagonal Fusion for Visual Question Answering and Visual Relationship Detection
Authors:Hedi Ben-younes, Rémi Cadene, Nicolas Thome, Matthieu Cord
[25] *Title: SCATGAN for Reconstruction of Ultrasound Scatterers Using Generative Adversarial Networks
Authors:Andrawes Al Bahou, Christine Tanner, Orcun Goksel
[26] Title: Scalable Learning-Based Sampling Optimization for Compressive Dynamic MRI
Authors:Thomas Sanchez, Baran Gözcü, Ruud B. van Heeswijk, Efe Ilıcak, Tolga Çukur, and Volkan Cevher
[27] Title: Natural and Adversarial Error Detection using Invariance to Image Transformations
Authors:Yuval Bahat, Michal Irani, Gregory Shakhnarovich
