mlsanityData to 7 Oct 2026

How much real data do we actually need: Analyzing object detection performance using synthetic and real data

arXiv 1907.07061 ↗First posted 16 Jul 2019cs.CVcited by 29

Not on the map. Why?

Abstract (arXiv)

In recent years, deep learning models have resulted in a huge amount of progress in various areas, including computer vision. By nature, the supervised training of deep models requires a large amount of data to be available. This ideal case is usually not tractable as the data annotation is a tremendously exhausting and costly task to perform. An alternative is to use synthetic data. In this paper, we take a comprehensive look into the effects of replacing real data with synthetic data. We further analyze the effects of having a limited amount of real data. We use multiple synthetic and real datasets along with a simulation tool to create large amounts of cheaply annotated synthetic data. We analyze the domain similarity of each of these datasets. We provide insights about designing a methodological procedure for training deep networks using these datasets.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). First posted on arXiv 16 Jul 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 3; 2022: 5; 2023: 8; 2024: 5; 2025: 7; 2026: 1.

29 papers.

Sort:

By area:

  1. 2023Radar-Camera Fusion for Object Detection and Semantic Segmentation in Autonomous Driving: A Comprehensive Reviewcited by 57
  2. 2024Efficacy of Synthetic Data as a Benchmarkcited by 29
  3. 2021Sim2Air - Synthetic aerial dataset for UAV monitoringcited by 16
  4. 2021VersatileGait: A Large-Scale Synthetic Gait Dataset with Fine-GrainedAttributes and Complicated Scenarioscited by 12
  5. 2024Advancing Fine-Grained Classification by Structure and Subject Preserving Augmentationcited by 12
  6. 2022Analysis of Training Object Detection Models with Synthetic Datacited by 9
  7. 2022StorSeismic: A new paradigm in deep learning for seismic processingcited by 7
  8. 2023Quantifying the LiDAR Sim-to-Real Domain Shift: A Detailed Investigation Using Object Detectors and Analyzing Point Clouds at Target-Levelcited by 7
  9. 2023Object Detector Differences when using Synthetic and Real Training Datacited by 5
  10. 2022Reducing the Amount of Real World Data for Object Detector Training with Synthetic Datacited by 4
  11. 2023Synthetic Data Generation for Bridging Sim2Real Gap in a Production Environmentcited by 4
  12. 2021Autoencoder Based Inter-Vehicle Generalization for In-Cabin Occupant Classificationcited by 3
17 more
  1. 2025Synthetic Industrial Object Detection: GenAI vs. Feature-Based Methodscited by 3
  2. 2022TRoVE: Transforming Road Scene Datasets into Photorealistic Virtual Environmentscited by 2
  3. 2023You Only Crash Once: Improved Object Detection for Real-Time, Sim-to-Real Hazardous Terrain Detection and Classification for Autonomous Planetary Landingscited by 2
  4. 2023Data and Knowledge for Overtaking Scenarios in Autonomous Drivingcited by 2
  5. 2024Bayesian Detector Combination for Object Detection with Crowdsourced Annotationscited by 2
  6. 2024Facing Asymmetry -- Uncovering the Causal Link between Facial Symmetry and Expression Classifiers using Synthetic Interventionscited by 2
  7. 2025Development of Hybrid Artificial Intelligence Training on Real and Synthetic Data: Benchmark on Two Mixed Training Strategiescited by 2
  8. 2023Realistically distributing object placements in synthetic training data improves the performance of vision-based object detection modelscited by 1
  9. 2023Deep Aramaic: Towards a Synthetic Data Paradigm Enabling Machine Learning in Epigraphycited by 1
  10. 2025Is AI currently capable of identifying wild oysters? A comparison of human annotators against the AI model, ODYSSEEcited by 1
  11. 2025Synthetic Dataset Generation for Autonomous Mobile Robots Using 3D Gaussian Splatting for Vision Trainingcited by 1
  12. 2025Can We Challenge Open-Vocabulary Object Detectors with Generated Content in Street Scenes?cited by 1
  13. 2026Does Synthetic Layered Design Data Benefit Layered Design Decomposition?cited by 1
  14. 2022How many radiographs are needed to re-train a deep learning system for object detection?cited by 0
  15. 2024OSSA: Unsupervised One-Shot Style Adaptationcited by 0
  16. 2025UrbanTwin: Building High-Fidelity Digital Twins for Sim2Real LiDAR Perception and Evaluationcited by 0
  17. 2025High-Fidelity Digital Twins for Bridging the Sim2Real Gap in LiDAR-Based ITS Perceptioncited by 0