Data to 7 Oct 2026

Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection

arXiv 1708.01642 ↗First posted 4 Aug 2017cs.CVcited by 209

Not on the map. Why?

Abstract (arXiv)

A major impediment in rapidly deploying object detection models for instance detection is the lack of large annotated datasets. For example, finding a large labeled dataset containing instances in a particular kitchen is unlikely. Each new environment with new instances requires expensive data collection and annotation. In this paper, we propose a simple approach to generate large annotated instance datasets with minimal effort. Our key insight is that ensuring only patch-level realism provides enough training signal for current object detector models. We automatically `cut' object instances and `paste' them on random backgrounds. A naive way to do this results in pixel artifacts which result in poor performance for trained models. We show how to make detectors ignore these artifacts during training and generate data that gives competitive performance on real data. Our method outperforms existing synthesis approaches and when combined with real images improves relative performance by more than 21% on benchmark datasets. In a cross-domain setting, our synthetic data combined with just 10% real data outperforms models trained on all real data.

Key facts. Cited by 209 papers in the map (data through 7 Oct 2026). First posted on arXiv 4 Aug 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 47; 2022: 52; 2023: 41; 2024: 32; 2025: 20; 2026: 17.

209 papers.

Sort:

By area:

2023Cut and Learn for Unsupervised Object Detection and Instance Segmentationcited by 186
2023TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Compositioncited by 182
2021Occluded Video Instance Segmentation: A Benchmarkcited by 154
2021Mix3D: Out-of-Context Data Augmentation for 3D Scenescited by 126
2021Pre-training without Natural Imagescited by 100
2021Instance Localization for Self-supervised Detection Pretrainingcited by 99
2022A Comprehensive Survey of Image Augmentation Techniques for Deep Learningcited by 95
2021GeoSim: Realistic Video Simulation via Geometry-Aware Composition for Self-Drivingcited by 75
2021Region-aware Adaptive Instance Normalization for Image Harmonizationcited by 71
2023BOP Challenge 2022 on Detection, Segmentation and Pose Estimation of Specific Rigid Objectscited by 70
2021Context Decoupling Augmentation for Weakly Supervised Semantic Segmentationcited by 69
2021Making Images Real Again: A Comprehensive Survey on Deep Image Compositioncited by 63

38 more

2021Randomness In Neural Network Training: Characterizing The Impact of Toolingcited by 57
2022X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusioncited by 49
2021DSP: Dual Soft-Paste for Unsupervised Domain Adaptive Semantic Segmentationcited by 45
2021Motion-aware Contrastive Video Representation Learning via Foreground-background Mergingcited by 45
2021Product1M: Towards Weakly Supervised Instance-Level Product Retrieval via Cross-modal Pretrainingcited by 42
2021Learning to Detect Every Thing in an Open Worldcited by 37
2022Neural Light Field Estimation for Street Scenes with Differentiable Virtual Object Insertioncited by 35
2023MosaicFusion: Diffusion Models as Data Augmenters for Large Vocabulary Instance Segmentationcited by 35
2022Occlusion-Aware Self-Supervised Monocular 6D Object Pose Estimationcited by 34
2024GeoDiffuser: Geometry-Based Image Editing with Diffusion Modelscited by 34
2024AeroGen: Enhancing Remote Sensing Object Detection with Diffusion-Driven Data Generationcited by 32
2022Unsupervised Pre-training for Temporal Action Localization Taskscited by 31
2021CarveMix: A Simple Data Augmentation Method for Brain Lesion Segmentationcited by 30
2022Neural-Sim: Learning to Generate Training Data with NeRFcited by 27
2021ZePHyR: Zero-shot Pose Hypothesis Ratingcited by 26
2021Occlusion Handling in Generic Object Detection: A Reviewcited by 25
2021Survey: Image Mixing and Deleting for Data Augmentationcited by 25
2022AcroFOD: An Adaptive Method for Cross-domain Few-shot Object Detectioncited by 25
2023VISION Datasets: A Benchmark for Vision-based InduStrial InspectiONcited by 25
2022SCS-Co: Self-Consistent Style Contrastive Learning for Image Harmonizationcited by 24
2021Image Compositing for Segmentation of Surgical Tools without Manual Annotationscited by 23
2021Scale-aware Automatic Augmentation for Object Detectioncited by 23
2022TexPose: Neural Texture Learning for Self-Supervised 6D Object Pose Estimationcited by 23
2022You Only Cut Once: Boosting Data Augmentation with a Single Cutcited by 22
2022Referring Image Mattingcited by 22
2024PrimeComposer: Faster Progressively Combined Diffusion for Image Composition with Attention Steeringcited by 20
2022InsMix: Towards Realistic Generative Data Augmentation for Nuclei Instance Segmentationcited by 19
2021Interpreting the Robustness of Neural NLP Models to Textual Perturbationscited by 17
2022DALL-E for Detection: Language-driven Compositional Image Synthesis for Object Detectioncited by 17
2023MoCaE: Mixture of Calibrated Experts Significantly Improves Object Detectioncited by 17
2023Instance Tracking in 3D Scenes from Egocentric Videoscited by 17
2024CuVLER: Enhanced Unsupervised Object Discoveries through Exhaustive Self-Supervised Transformerscited by 16
2023MMNet: Multi-Collaboration and Multi-Supervision Network for Sequential Deepfake Detectioncited by 15
20233D Copy-Paste: Physically Plausible Object Insertion for Monocular 3D Detectioncited by 15
2021A Random CNN Sees Objects: One Inductive Bias of CNN and Its Applicationscited by 14
2023IPMix: Label-Preserving Data Augmentation Method for Training Robust Classifierscited by 14
2024Scaling Backwards: Minimal Synthetic Pre-training?cited by 14
2024Past, Present, and Future of Sensor-Based Human Activity Recognition Using Wearables: A Surveying Tutorial on a Still Challenging Taskcited by 14

50 most cited of 209 All 209 (JSON)