YouTube-VOS: A Large-Scale Video Object Segmentation Benchmark
arXiv:1809.03327
Abstract
Learning long-term spatial-temporal features are critical for many video analysis tasks. However, existing video segmentation methods predominantly rely on static image segmentation techniques, and methods capturing temporal dependency for segmentation have to depend on pretrained optical flow models, leading to suboptimal solutions for the problem. End-to-end sequential learning to explore spatialtemporal features for video segmentation is largely limited by the scale of available video segmentation datasets, i.e., even the largest video segmentation dataset only contains 90 short video clips. To solve this problem, we build a new large-scale video object segmentation dataset called YouTube Video Object Segmentation dataset (YouTube-VOS). Our dataset contains 4,453 YouTube video clips and 94 object categories. This is by far the largest video object segmentation dataset to our knowledge and has been released at http://youtube-vos.org. We further evaluate several existing state-of-the-art video object segmentation algorithms on this dataset which aims to establish baselines for the development of new algorithms in the future.
arXiv commentDataset Report. arXiv admin note: substantial text overlap with arXiv:1809.00461
References in corpus (4)
Cited by in corpus (59)
- Deep Learning for Visual Tracking: A Comprehensive Survey
- BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning
- Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation
- Cell Segmentation and Tracking using CNN-Based Distance Predictions and a Graph-Based Matching Strategy
- SkyNet: a Hardware-Efficient Method for Object Detection and Tracking on Embedded Systems
- Video Object Segmentation with Adaptive Feature Bank and Uncertain-Region Refinement
- Decoupled Spatial-Temporal Transformer for Video Inpainting
- A Comprehensive Review of Modern Object Segmentation Approaches
- Suppress and Balance: A Simple Gated Network for Salient Object Detection
- RVOS: End-to-End Recurrent Network for Video Object Segmentation
- Efficient Long-Short Temporal Attention Network for Unsupervised Video Object Segmentation
- Adversarial Attacks on Video Object Segmentation with Hard Region Discovery
- Collaborative Video Object Segmentation by Foreground-Background Integration
- Associating Objects with Transformers for Video Object Segmentation
- HODOR: High-level Object Descriptors for Object Re-segmentation in Video Learned from Static Images
- Fully Transformer-Equipped Architecture for End-to-End Referring Video Object Segmentation
- Learning Joint Spatial-Temporal Transformations for Video Inpainting
- Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding
- CompFeat: Comprehensive Feature Aggregation for Video Instance Segmentation
- Video Instance Segmentation
- DMM-Net: Differentiable Mask-Matching Network for Video Object Segmentation
- Modular Interactive Video Object Segmentation: Interaction-to-Mask, Propagation and Difference-Aware Fusion
- Collaborative Video Object Segmentation by Multi-Scale Foreground-Background Integration
- Betrayed by Motion: Camouflaged Object Discovery via Motion Segmentation
- MAST: A Memory-Augmented Self-supervised Tracker
- FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting
- A Transductive Approach for Video Object Segmentation
- Fast video object segmentation with Spatio-Temporal GANs
- VORNet: Spatio-temporally Consistent Video Inpainting for Object Removal
- KinePose: A temporally optimized inverse kinematics technique for 6DOF human pose estimation with biomechanical constraints
- Self-supervised Video Object Segmentation by Motion Grouping
- RiWNet: A moving object instance segmentation Network being Robust in adverse Weather conditions
- TTVOS: Lightweight Video Object Segmentation with Adaptive Template Attention Module and Temporal Consistency Loss
- Deep Flow-Guided Video Inpainting
- MouseSIS: A Frames-and-Events Dataset for Space-Time Instance Segmentation of Mice
- One-Shot Weakly Supervised Video Object Segmentation
- Deep Learning for Robust Motion Segmentation with Non-Static Cameras
- Learning to Track Instances without Video Annotations
- U2-ONet: A Two-level Nested Octave U-structure with Multiscale Attention Mechanism for Moving Instances Segmentation
- PMVOS: Pixel-Level Matching-Based Video Object Segmentation
- CRVOS: Clue Refining Network for Video Object Segmentation
- Discriminative Online Learning for Fast Video Object Segmentation
- Learning Spatio-Appearance Memory Network for High-Performance Visual Tracking
- Fast Video Object Segmentation With Temporal Aggregation Network and Dynamic Template Matching
- An Exploration of Target-Conditioned Segmentation Methods for Visual Object Trackers
- Self-supervised Object Tracking with Cycle-consistent Siamese Networks
- Cross-Modal Progressive Comprehension for Referring Segmentation
- Two stages for visual object tracking
- Learning Dynamic Network Using a Reuse Gate Function in Semi-supervised Video Object Segmentation
- Guided Interactive Video Object Segmentation Using Reliability-Based Attention Maps
- Dual Temporal Memory Network for Efficient Video Object Segmentation
- An Empirical Study of Propagation-based Methods for Video Object Segmentation
- DyStaB: Unsupervised Object Segmentation via Dynamic-Static Bootstrapping
- Learning Dynamic Compact Memory Embedding for Deformable Visual Object Tracking
- DynOcc: Learning Single-View Depth from Dynamic Occlusion Cues
- RPM-Net: Robust Pixel-Level Matching Networks for Self-Supervised Video Object Segmentation
- VideoClick: Video Object Segmentation with a Single Click
- Interactive Video Object Segmentation Using Global and Local Transfer Modules
- Partially Observed Exchangeable Modeling