2019顶会论文整理-MOT多目标跟踪

---

我目前主要研究方向是车辆的3D多目标跟踪(3D MOT)。在2019年计算机视觉顶会ICCV和CVPR中整理了相关论文，以下有video链接的为oral。有源码的已经列出方便复现，没有源码就…见仁见智～摊手～

---

new idea

hightlight	Title	Org	code
TBA	Tracking by Animation- Unsupervised Learning of Multi-Object Attentive Trackers	CVPR	code
converting a detector into a Tracktor (SOTA)	Tracking without bells and whistles	ICCV	code

unify network

hightlight	Title	Org	code
a unified framework for similarity measurement	Spatial-Temporal Relation Networks for Multi-Object Tracking	ICCV	-
MOT: joint learning, end-to-end	FAMNet: Joint Learning of Feature, Affinity and Multi-dimensional Assignment for Online Multiple Object Tracking [Video]	ICCV	-
Unify detection&tracking	Bridging the Gap Between Detection and Tracking: A Unified Approach	ICCV	-
多模态融合，end-to-end，第一个对点云提取深度特征用于association 把我的idea给做了TAT	Robust Multi-Modality Multi-Object Tracking	ICCV	code

point cloud

hightlight	Title	Org	code
3D Siamese tracker applied to point clouds	Leveraging Shape Completion for 3D Siamese Tracking	CVPR	code
new conv: fast to train (SOTA)	ShellNet: Efficient Point Cloud Convolutional Neural Networks Using Concentric Shells Statistics [Video]	ICCV	code
new conv: feature learning	Interpolated Convolutional Networks for 3D Point Cloud Understanding [Video]	ICCV	-
点云特征图	PointCloud Saliency Maps [Video]	ICCV	code