Object Tracking

目录

ECCV 2022

ECCV 2020

ICCV 2023

CVPR 2023

CVPR 2022

---

ECCV 2022

1.（MOT、指标）MOTCOM: The Multi-Object Tracking Dataset Complexity Metric

2.（鱼数据集、声呐视频、MOT）The Caltech Fish Counting Dataset: A Benchmark for Multiple-Object Tracking and Counting

3.（MOT多人、数据集）Large Scale Real-World Multi-person Tracking 

4.（细胞跟踪）Graph Neural Network for Cell Tracking in Microscopy Videos 图神经网络方法

5.（SOT与MOT）Towards Grand Unification of Object Tracking 使用相同的模型参数通过单个网络同时解决四个跟踪问题（SOT、MOT、VOS、MOTS）

6.（MOT）ByteTrack: Multi-Object Tracking by Associating Every Detection Box 提出一种简单有效数据关联-data association方法

7（MOT）.Robust Multi-Object Tracking by Marginal Inference 

8.（MOT、3d）PolarMOT: How Far Can Geometric Relations Take Us in 3D Multi-Object Tracking? 仅使用几何关系进行数据关联、图神经网络方法

9.Particle Video Revisited: Tracking through Occlusions Using Point Trajectories 点轨迹光流估计

10.（MOT）Tracking Objects As Pixel-Wise Distributions 像素分布跟踪，transformer网络

11.（RGBD相机、数据集、3d）Towards Generic 3D Tracking in RGBD Videos: Benchmark and Baseline 

12.AiATrack: Attention in Attention for Transformer Visual Tracking 注意力机制、transformer网络

13.（医学）Robust Landmark-Based Stent Tracking in X-Ray Fluoroscopy 图卷积神经网络

14.（VOT）Backbone Is All Your Need: A Simplified Architecture for Visual Object Tracking

15.Hierarchical Feature Embedding for Visual Tracking 分层特征嵌入模型

16.（MOT）Tracking Every Thing in the Wild 引入新的指标跟踪每件事的准确性（TETA）

17.Towards Sequence-Level Training for Visual Tracking 基于强化学习的视觉跟踪序列级训练

18.（分割）Robust Visual Tracking by Segmentation 

19.（3d）Large-Displacement 3D Object Tracking with Hybrid Non-local Optimization 针对不同参数结合非局部和局部优化的混合方法

20."FEAR: Fast, Efficient, Accurate and Robust Visual Tracker" 

21.（平面跟踪）"HVC-Net: Unifying Homography, Visibility, and Confidence Learning for Planar Object Tracking" 

22.Tracking by Associating Clips 将单个长视频序列视为多个短剪辑，然后在剪辑内和剪辑之间执行跟踪

23.（MOT、建模）MOTR: End-to-End Multiple-Object Tracking with TRansformer transformer

24.Bayesian Tracking of Video Graphs Using Joint Kalman Smoothing and Registration 开发了一种卡尔曼平滑方法，用于从噪声、杂乱和不完整的数据中估计图形 

25.（3d、交通场景）SpOT: Spatiotemporal Modeling for 3D Object Tracking 将跟踪对象表示为时间戳点和边界框的序列，将跟踪重新表述为时空问题，利用场景中参与者的空间和时间信息

26.（人）"BodySLAM: Joint Camera Localisation, Mapping, and Human Motion Tracking" 

27.（人）AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion Sensing 

28.（3d、lidar）3D Siamese Transformer Network for Single Object Tracking on Point Clouds transformer网络

ECCV 2020

1.Segment as Points for Efficient Online Multi-Object Tracking and Segmentation 

2.Tracking Objects as Points 

3.TAO: A Large-Scale Benchmark for Tracking Any Object 

4.Towards Real-Time Multi-Object Tracking 

5.Object Tracking using Spatio-Temporal Networks for Future Prediction Location  

6.Challenge-Aware RGBT Tracking 

7.PG-Net: Pixel to Global Matching Network for Visual Tracking 

8.Know Your Surroundings: Exploiting Scene Information for Object Tracking 

9.Simultaneous Detection and Tracking with Motion Modelling for Multiple Object Tracking 

10.Tracking Emerges by Looking Around Static Scenes, with Neural 3D Mapping 

ICCV 2023

1.Tracking by 3D Model Estimation of Unknown Objects in Videos 

2.TrackFlow: Multi-Object tracking with Normalizing Flows 

3.A Fast Unified System for 3D Object Detection and Tracking 

4.MixCycle: Mixup Assisted Semi-Supervised 3D Single Object Tracking with Cycle Consistency 

5.Foreground-Background Distribution Modeling Transformer for Visual Object Tracking 

6.Tracking Anything with Decoupled Video Segmentation 

7.Heterogeneous Diversity Driven Active Learning for Multi-Object Tracking 

8.Humans in 4D: Reconstructing and Tracking Humans with Transformers 

9.Delving into Motion-Aware Matching for Monocular 3D Object Tracking 

10.（物体建模）Robust Object Modeling for Visual Tracking 

11.Integrating Boxes and Masks: A Multi-Object Framework for Unified Visual Tracking and Segmentation12.3DMOTFormer: Graph Transformer for Online 3D Multi-Object Tracking 

13.Object-Centric Multiple Object Tracking 

14.TrajectoryFormer: 3D Object Tracking Transformer with Predictive Trajectory Hypotheses 

15.Tracking Everything Everywhere All at Once 

16.（数据集）！！SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports Scenes 

17.Uncertainty-aware Unsupervised Multi-Object Tracking 

18.（无人机）Adaptive and Background-Aware Vision Transformer for Real-Time UAV Tracking 

19.！！ Multiple Planar Object Tracking

20.（数据集）360VOT: A New Benchmark Dataset for Omnidirectional Visual Object Tracking 

21.Tracking without Label: Unsupervised Multiple Object Tracking via Contrastive Similarity Learning22.TAPIR: Tracking Any Point with Per-Frame Initialization and Temporal Refinement

23.（人体）TEMPO: Efficient Multi-View Pose Estimation, Tracking, and Forecasting 

24.Synchronize Feature Extracting and Matching: A Single Branch Framework for 3D Object Tracking25.TAPIR: Tracking Any Point with Per-Frame Initialization and Temporal Refinement 

26.Collaborative Tracking Learning for Frame-Rate-Insensitive Multi-Object Tracking 

27.（数据集）PlanarTrack: A Large-scale Challenging Benchmark for Planar Object Tracking 

28.Deep Active Contours for Real-time 6-DoF Object Tracking 

29.（联合文本）CiteTracker: Correlating Image and Text for Visual Tracking 

30.（人体）Human from Blur: Human Pose Tracking from Blurry Images 

CVPR 2023

1.（无人机）Resource-Efficient RGBD Aerial Tracking 

2.（视觉表征）Representation Learning for Visual Object Tracking by Masked Appearance Transfer 

3.（序列回归）SeqTrack: Sequence to Sequence Learning for Visual Object Tracking 

4.Frame-Event Alignment and Fusion Network for High Frame Rate Tracking 融合传统帧和基于事件相机

5.OVTrack: Open-Vocabulary Multiple Object Tracking 预训练之外类别的鲁棒性、大规模现实场景

6.MotionTrack: Learning Robust Short-Term and Long-Term Motions for Multi-Object Tracking

7.Bridging Search Region Interaction With Template for RGB-T Tracking

8.（密集遮挡）MotionTrack: Learning Robust Short-Term and Long-Term Motions for Multi-Object Tracking 

9.（lidar）VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Tracking

10.（多摄像头）Standing Between Past and Future: Spatio-Temporal Modeling for Multi-Camera 3D Multi-Object Tracking

11.Focus on Details: Online Multi-Object Tracking With Diverse Fine-Grained Representation 细节跟踪

12.（细胞）Unsupervised Contour Tracking of Live Cells by Mechanical and Cycle Consistency Losses

13.（虚拟VR服装）GarmentTracking: Category-Level Garment Pose Tracking

14.VideoTrack: Learning To Track Objects via Video Transformer

15.（lidar、数据集）ARKitTrack: A New Diverse Dataset for Tracking Using Mobile RGB-D Data

16.UTM: A Unified Multiple Object Tracking Model With Identity-Aware Feature Enhancement 

17.Unifying Short and Long-Term Tracking With Graph Hierarchies 长视频物体跟踪

18.（互动跟踪、重建）Visibility Aware Human-Object Interaction Tracking From Single RGB Camera

19.（联合文本定位）Joint Visual Grounding and Tracking With Natural Language Specification

20.Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking 修正短时累计误差

21.（多模态跟踪）Visual Prompt Multi-Modal Tracking

22.（重建）BundleSDF: Neural 6-DoF Tracking and 3D Reconstruction of Unknown Objects

23.(数据集、动物轨迹跟踪）3D-POP - An Automated Annotation Approach to Facilitate Markerless 2D-3D Tracking of Freely Moving Birds With Marker-Based Motion Capture 用运动捕捉（mo-cap）系统，以半自动方式获取大量动物运动和姿势（二维和三维）注释数据

24.（事件摄像机跟踪器） Data-Driven Feature Tracking for Event Cameras

25.（数据集）JRDB-Pose: A Large-Scale Dataset for Multi-Person Pose Estimation and Tracking

26.（联合文本语言）Referring Multi-Object Tracking

27.Autoregressive Visual Tracking

28. （虚拟）Avatars Grow Legs: Generating Smooth Human Motion From Sparse Tracking Inputs With Diffusion Model

29.MOTRv2: Bootstrapping End-to-End Multi-Object Tracking by Pretrained Object Detectors 

CVPR 2022

1.（无纹理）
Iterative Corresponding Geometry: Fusing Region and Depth for Highly Efficient 3D Tracking of Textureless Objects概率跟踪器 ICG，它融合了区域和深度信息，只需要物体的几何形状

2. （开放世界）Opening Up Open World Tracking 阐述了在开放世界环境中检测和跟踪已知和未知物体的前景和评估方法，提出OWTB方法应对。

3.（UAV）Multi-Object Tracking Meets Moving UAV 

4.（人物交互、数据集）BEHAVE: Dataset and Method for Tracking Human Object Interactions

5.（人）Tracking People by Predicting 3D Appearance, Location and Pose

6.Transforming Model Prediction for Tracking

7.（长时间跨度）MeMOT: Multi-Object Tracking With Memory 保留一个大型时空存储器来存储跟踪对象的身份嵌入

8. (动物行为分析）PyMiceTracking: An Open-Source Toolbox for Real-Time Behavioral Neuroscience Experiments

9.（视频跟踪）Efficient Video Instance Segmentation via Tracklet Query and Proposal 

小轨迹查询，端到端的视频实例分类、分割、跟踪 

10. （全局多目标跟踪）Global Tracking Transformers

11.（数据集、真实3d、SOT）BCOT: A Markerless High-Precision 3D Object Tracking Benchmark 无需标记，使用双目数据估计高精度真实姿态，面向单目无纹理三维物体

12.（无监督）Unsupervised Learning of Accurate Siamese Tracking

13.（SOT与MOT均可）Unified Transformer Tracker for Object Tracking

14.（SOT、事件摄像机）Spiking Transformers for Event-Based Single Object Tracking 动态提取和融合时域和空间域的信息

15.（夜间航空跟踪、数据集）Unsupervised Domain Adaptation for Nighttime Aerial Tracking

16.（UAV）TCTrack: Temporal Contexts for Aerial Tracking 利用时间信息增强空间特征

17.（自动驾驶、轨迹预测）Whose Track Is It Anyway? Improving Robustness to Tracking Errors With Affinity-Based Trajectory Prediction

18.（自动驾驶、单目视频）Time3D: End-to-End Joint Monocular 3D Object Detection and Tracking for Autonomous Driving

19.（MOT） Adiabatic Quantum Computing for Multi Object Tracking 

20.（在线、MOT）Towards Discriminative Representation: Multi-View Trajectory Contrastive Learning for Online Multi-Object Tracking 多视角轨迹对比学习

21.（多相机MOT）LMGP: Lifted Multicut Meets Geometry Projections for Multi-Camera Multi-Object Tracking

22.（手部、视频数据集）

Forward Propagation, Backward Regression, and Pose Association for Hand Tracking in the Wild 卷积结构实现大规模无约束视频标注手的位置以及轨迹

23.Correlation-Aware Deep Tracking

24.Transformer Tracking With Cyclic Shifting Window Attention 多尺度循环移窗注意力的新型变换器架构

25.（长期跟踪）Global Tracking via Ensemble of Local Trackers 结合两种长期跟踪策略，多跟踪器进行跟踪

26.MixFormer: End-to-End Tracking With Iterative Mixed Attention 混合注意力模块（MAM），用于同时提取特征和整合目标信息

27.（lidar）PTTR: Relational 3D Point Cloud Object Tracking With Transformer

28.（MOT、3d）TrackFormer: Multi-Object Tracking With Transformers 省去任何额外的图形优化或运动和/或外观建模

29.Ranking-Based Siamese Visual Tracking 提出了一种基于排序的优化算法

30.（lidar、3d）Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds 提出了一种无匹配的两阶段跟踪器，引入了以运动为中心的范式。

31. （MOT）Learning of Global Objective for Network Flow in Multi-Object Tracking 基于最小成本流（MCF）表述的多目标跟踪问题

32.（人、MOT、数据集）PoseTrack21: A Dataset for Person Search, Multi-Object Tracking and Multi-Person Pose Tracking

33. （可见光温度RGBT、UAV、数据集）Visible-Thermal UAV Tracking: A Large-Scale Benchmark and New Baseline 

34.（多人跟踪、MOT、数据集）DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse Motion