今日CS.CV计算机视觉论文速览
Wed, 6 Feb 2019
Totally 14 papers
Interesting:
-
AVP，adaptive virtual patients,一种用于小数据学习的基于物理的数据生成手段。利用adaptive virtual patients的方法，基于双边解耦合生成可逆网络来从病理数据中抽取特征，并合成具有病理学意义的虚拟病患数据。(from 佐治亚理工)
-
单阶段的人脸检测网络,基于resnet101实现的快速人脸检测，充分利用深层网络的上下文信息，在保持低假阳性的同时提高了召回率，并对输出进行编码减小了检测时间。（from 日本东北大学）
网络框架：
融合与检测模块：
dataset:
WIDER FACE
FDDB: Face Detection Data Set and Benchmark
Multi-Attribute Labelled Faces, MALF
IARPA
Labeled Faces in the Wild
Face Detection Performance
ref:
pascal obj:http://host.robots.ox.ac.uk/pascal/VOC/databases.html
pascal:http://host.robots.ox.ac.uk/pascal/VOC/ -
TrackNet, 同时实现目标检测与跟踪的深度学习模型。研究人员基于fast RCNN进行改进、实现了视频中目标的移动目标三维检测。由于在连续帧中目标运动呈现一个三维管道，文中还提出了Tube推举网络来预测物体的运动。这个网络可以对目标进行直接检测和跟踪。（from 纽约大学）
文章中提出的网络架构：
TPN网络的具体结构：
检测与跟踪的效果：
数据集：户外交通数据UADETRAC -
EasyLabel, 一个用于创建机器人RGB-D的数据集的半自动标注工具。这一工具可以半自动化的生成高精度的目标掩膜，研究人员利用它合成了Object Cluttered Indoor Dataset(OCID)。(from 奥地利维也纳技术大学)注意参考文献
实验装置设计：
一些数据集比较效果：
数据集：https://www.acin.tuwien.ac.at/object-clutter-indoor-dataset/
ref:rgbd相机的综合测评](https://ieeexplore.ieee.org/document/8436009) -
基于生成对抗进行火焰检测, 为了对包含火焰/序列和时序信息有鲁棒的表示，利用DCGAN来生成火焰并检测，克服了标记数据和时序信息缺失的限制。
网络架构如下图所示：
检测结果：
Daily Computer Vision Papers
[1] Title: Face Alignment using a 3D Deeply-initialized Ensemble of Regression Trees
Authors:Roberto Valle (1), José M. Buenaposada (2), Antonio Valdés (3), Luis Baumela (1) ((1) Universidad Politécnica de Madrid, (2) Universidad Rey Juan Carlos, (3) Universidad Complutense de Madrid)
[2] Title: Deep Convolutional Generative Adversarial Networks Based Flame Detection in Video
Authors:Süleyman Aslan, Uğur Güdükbay, B. Uğur Töreyin, A. Enis Çetin
[3] Title: An RNN-based IMM Filter Surrogate
Authors:Stefan Becker, Ronny Hug, Wolfgang Hübner, Michael Arens
[4] Title: 6D Object Pose Estimation without PnP
Authors:Jin Liu
[5] Title: DVOLVER: Efficient Pareto-Optimal Neural Network Architecture Search
Authors:Guillaume Michel, Mohammed Amine Alaoui, Alice Lebois, Amal Feriani, Mehdi Felhi
[6] Title: Revisiting a single-stage method for face detection
Authors:Nguyen Van Quang, Hiromasa Fujihara
[7] Title: SASSE: Scalable and Adaptable 6-DOF Pose Estimation
Authors:Huu Le, Tuan Hoang, Qianggong Zhang, Thanh-Toan Do, Anders Eriksson, Michael Milford
[8] Title: A Practical Maximum Clique Algorithm for Matching with Pairwise Constraints
Authors:Alvaro Parra Bustos, Tat-Jun Chin, Frank Neumann, Tobias Friedrich, Maximilian Katzmann
[9] Title: AVP: Physics-informed Data Generation for Small-data Learning
Authors:Jialei Chen, Yujia Xie, Kan Wang, Chuck Zhang, Mani A. Vannan, Ben Wang, Zhen Qian
[10] Title: A Two-Stream Siamese Neural Network for Vehicle Re-Identification by Using Non-Overlapping Cameras
Authors:Icaro O. de Oliveira, Keiko V. O. Fonseca, Rodrigo Minetto
[11] Title: TrackNet: Simultaneous Object Detection and Tracking and Its Application in Traffic Video Analysis
Authors:Chenge Li, Gregory Dobler, Xin Feng, Yao Wang
[12] Title: Very Long Term Field of View Prediction for 360-degree Video Streaming
Authors:Chenge Li, Weixi Zhang, Yong Liu, Yao Wang
[13] Title: Partial Fingerprint Detection Using Core Point Location
Authors:Wajih Ullah Baig, Adeel Ejaz, Umar Munir, Kashif Sardar
[14] Title: EasyLabel: A Semi-Automatic Pixel-wise Object Annotation Tool for Creating Robotic RGB-D Datasets
Authors:Markus Suchi, Timothy Patten, Markus Vincze
