GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild
arXiv:1810.11981 · doi:10.1109/TPAMI.2019.2957464
Abstract
We introduce here a large tracking database that offers an unprecedentedly wide coverage of common moving objects in the wild, called GOT-10k. Specifically, GOT-10k is built upon the backbone of WordNet structure and it populates the majority of over 560 classes of moving objects and 87 motion patterns, magnitudes wider than the most recent similar-scale counterparts. The contributions of this paper are summarized in the following: (1) GOT-10k offers over 10,000 video segments with more than 1.5 million manually labeled bounding boxes, enabling unified training and stable evaluation of deep trackers. (2) GOT-10k is by far the first video trajectory dataset that uses the semantic hierarchy of WordNet to guide class population. (3) For the first time, GOT-10k introduces the one-shot protocol for tracker evaluation, where the training and test classes are zero-overlapped. The protocol avoids biased evaluation results towards familiar objects and it promotes generalization in tracker development. (4) We conduct extensive tracking experiments with 39 typical tracking algorithms on GOT-10k and analyze their results in this paper. (5) Finally, we develop a comprehensive platform for the tracking community that offers full-featured evaluation toolkits, an online evaluation server, and a responsive leaderboard. The annotations of GOT-10k's test data are kept private to avoid tuning parameters on it. The database, toolkits, evaluation server and baseline results are available at http://got-10k.aitestunion.com.
Accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
References in corpus (18)
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
- High-Speed Tracking with Kernelized Correlation Filters
- The Kinetics Human Action Video Dataset
- Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking
- Learning Spatially Regularized Correlation Filters for Visual Tracking
- MOT16: A Benchmark for Multi-Object Tracking
- Matching Networks for One Shot Learning
- MOTChallenge 2015: Towards a Benchmark for Multi-Target Tracking
- T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos
- The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scale
- Object Detection from Video Tubelets with Convolutional Neural Networks
- DCFNet: Discriminant Correlation Filters Network for Visual Tracking
- PTB-TIR: A Thermal Infrared Pedestrian Tracking Benchmark
- Visual object tracking performance measures revisited
- Siamese Instance Search for Tracking
- End-to-end Flow Correlation Tracking with Spatial-temporal Attention
- LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking
- An In-Depth Analysis of Visual Tracking with Siamese Neural Networks
Cited by in corpus (120)
- Deep Learning for Visual Tracking: A Comprehensive Survey
- Demystifying Contrastive Self-Supervised Learning: Invariances, Augmentations and Dataset Biases
- SkyNet: a Hardware-Efficient Method for Object Detection and Tracking on Embedded Systems
- Transformers in Single Object Tracking: An Experimental Survey
- UAVs Meet LLMs: Overviews and Perspectives Toward Agentic Low-Altitude Mobility
- Anti-UAV: A Large Multi-Modal Benchmark for UAV Tracking
- SiamFC++: Towards Robust and Accurate Visual Tracking with Target Estimation Guidelines
- Semi-Supervised Visual Tracking of Marine Animals using Autonomous Underwater Vehicles
- Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking
- Adaptive Siamese Tracking with a Compact Latent Network
- Visual Object Tracking in First Person Vision
- Siamese Infrared and Visible Light Fusion Network for RGB-T Tracking
- SiamCAR: Siamese Fully Convolutional Classification and Regression for Visual Tracking
- Accurate Bounding-box Regression with Distance-IoU Loss for Visual Tracking
- Watching the World Go By: Representation Learning from Unlabeled Videos
- Target Transformed Regression for Accurate Tracking
- Unsupervised Cross-Modal Distillation for Thermal Infrared Tracking
- Beyond Greedy Search: Tracking by Multi-Agent Reinforcement Learning-based Beam Search
- TRAT: Tracking by Attention Using Spatio-Temporal Features
- DashNet: A Hybrid Artificial and Spiking Neural Network for High-speed Object Tracking
- BACTrack: Building Appearance Collection for Aerial Tracking
- Rethinking of Pedestrian Attribute Recognition: A Reliable Evaluation under Zero-Shot Pedestrian Identity Setting
- Deep Reinforcement Learning in Computer Vision: A Comprehensive Survey
- LaSOT: A High-quality Benchmark for Large-scale Single Object Tracking
- Global Instance Tracking: Locating Target More Like Humans
- In Defense and Revival of Bayesian Filtering for Thermal Infrared Object Tracking
- VisEvent: Reliable Object Tracking via Collaboration of Frame and Event Flows
- Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation
- Weakly-Supervised Domain Adaptation of Deep Regression Trackers via Reinforced Knowledge Distillation
- STMTrack: Template-free Visual Tracking with Space-time Memory Networks
- GlobalTrack: A Simple and Strong Baseline for Long-term Tracking
- Siamese Box Adaptive Network for Visual Tracking
- Exploring Dynamic Transformer for Efficient Object Tracking
- Ocean: Object-aware Anchor-free Tracking
- Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark
- BioDrone: A Bionic Drone-based Single Object Tracking Benchmark for Robust Vision
- How to Train Your Energy-Based Model for Regression
- Tracking by Instance Detection: A Meta-Learning Approach
- UncTrack: Reliable Visual Object Tracking with Uncertainty-Aware Prototype Memory Network
- Integration of the 3D Environment for UAV Onboard Visual Object Tracking
- RPT: Learning Point Set Representation for Siamese Visual Tracking
- BoLTVOS: Box-Level Tracking for Video Object Segmentation
- Tracking-by-Trackers with a Distilled and Reinforced Model
- Recursive Least-Squares Estimator-Aided Online Learning for Visual Tracking
- 1st Place Solution to ECCV-TAO-2020: Detect and Represent Any Object for Tracking
- LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search
- Spatio-temporal Prompting Network for Robust Video Feature Extraction
- Learn to Match: Automatic Matching Network Design for Visual Tracking
- Siamese Keypoint Prediction Network for Visual Object Tracking
- FMODetect: Robust Detection of Fast Moving Objects
- HiFT: Hierarchical Feature Transformer for Aerial Tracking
- SOTVerse: A User-defined Task Space of Single Object Tracking
- Real-Time Siamese Multiple Object Tracker with Enhanced Proposals
- Tracking Skiers from the Top to the Bottom
- State-Aware Tracker for Real-Time Video Object Segmentation
- Towards Effective and Efficient Adversarial Defense with Diffusion Models for Robust Visual Tracking
- TracKlinic: Diagnosis of Challenge Factors in Visual Tracking
- Effects of Blur and Deblurring to Visual Object Tracking
- SiamCorners: Siamese Corner Networks for Visual Tracking
- 3D Visual Tracking Framework with Deep Learning for Asteroid Exploration
- Learning Feature Embeddings for Discriminant Model based Tracking
- Cooling-Shrinking Attack: Blinding the Tracker with Imperceptible Noises
- Improving Visual Object Tracking through Visual Prompting
- RBCN: Rectified Binary Convolutional Networks for Enhancing the Performance of 1-bit DCNNs
- COMET: Context-Aware IoU-Guided Network for Small Object Tracking
- Learning to Filter: Siamese Relation Network for Robust Tracking
- Predictive Visual Tracking: A New Benchmark and Baseline Approach
- Faster and Simpler Siamese Network for Single Object Tracking
- Siamese Anchor Proposal Network for High-Speed Aerial Tracking
- Performance Evaluation Methodology for Long-Term Visual Object Tracking
- Deeply Activated Salient Region for Instance Search
- Domain Adaptive SiamRPN++ for Object Tracking in the Wild
- ROAM: Recurrently Optimizing Tracking Model
- Real-Time Visual Object Tracking via Few-Shot Learning
- Robust Visual Object Tracking with Two-Stream Residual Convolutional Networks
- CRACT: Cascaded Regression-Align-Classification for Robust Visual Tracking
- Tracking Without Re-recognition in Humans and Machines
- Accurate Anchor Free Tracking
- Towards real-time and energy efficient Siamese tracking -- a hardware-software approach
- An Exploration of Target-Conditioned Segmentation Methods for Visual Object Trackers
- Temporally-Transferable Perturbations: Efficient, One-Shot Adversarial Attacks for Online Visual Object Trackers
- Graph Attention Tracking
- Depth-only Object Tracking
- Pyramid Correlation based Deep Hough Voting for Visual Object Tracking
- AViTMP: A Tracking-Specific Transformer for Single-Branch Visual Tracking
- SiamRCR: Reciprocal Classification and Regression for Visual Object Tracking
- Object Localization Through a Single Multiple-Model Convolutional Neural Network with a Specific Training Approach
- Hard Negative Samples Emphasis Tracker without Anchors
- Learning Spatio-Appearance Memory Network for High-Performance Visual Tracking
- PatchNet -- Short-range Template Matching for Efficient Video Processing
- SiamSNN: Siamese Spiking Neural Networks for Energy-Efficient Object Tracking
- DAL -- A Deep Depth-aware Long-term Tracker
- Is First Person Vision Challenging for Object Tracking?
- Spatio-Temporal Matching for Siamese Visual Tracking
- Transparent Object Tracking Benchmark
- Efficient video annotation with visual interpolation and frame selection guidance
- Explicitly Modeling the Discriminability for Instance-Aware Visual Object Tracking
- Two stages for visual object tracking
- Implicit Equivariance in Convolutional Networks
- Unsupervised Visual Representation Learning by Tracking Patches in Video
- Cascaded Regression Tracking: Towards Online Hard Distractor Discrimination
- EXOT: Exit-aware Object Tracker for Safe Robotic Manipulation of Moving Object
- DMV: Visual Object Tracking via Part-level Dense Memory and Voting-based Retrieval
- Improving Human Annotation in Single Object Tracking
- Dynamic Attention guided Multi-Trajectory Analysis for Single Object Tracking
- An Analysis of Object Representations in Deep Visual Trackers
- DCF-ASN: Coarse-to-fine Real-time Visual Tracking via Discriminative Correlation Filter and Attentional Siamese Network
- IG-TRACK: IOU Guided Siamese Networks for visual object tracking
- A Simple and Strong Baseline for Universal Targeted Attacks on Siamese Visual Tracking
- Few-Shot Video Object Detection
- Tracking by Joint Local and Global Search: A Target-aware Attention based Approach
- Higher Performance Visual Tracking with Dual-Modal Localization
- Contrastive Learning of Image Representations with Cross-Video Cycle-Consistency
- Tracking the Untrackable
- SFTrack++: A Fast Learnable Spectral Segmentation Approach for Space-Time Consistent Tracking
- Effective Algorithm-Accelerator Co-design for AI Solutions on Edge Devices
- Saliency-Associated Object Tracking
- Leveraging Tacit Information Embedded in CNN Layers for Visual Tracking
- BOBBY2: Buffer Based Robust High-Speed Object Tracking
- Learning Dynamic Compact Memory Embedding for Deformable Visual Object Tracking