An Annotation Saved is an Annotation Earned: Using Fully Synthetic Training for Object Detection

Hinterstoißer, Stefan; Pauly, Olivier; Heibel, Hauke; Marek, Martina; Bokeloh, Martin

doi:10.1109/iccvw.2019.00340

Cited by 97 publications

(63 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Advancements in physics engines and graphics processing have advanced AI environment and data-generation capabilities, putting increased emphasis on transitioning models across the simulation-to-reality gap [21][22][23] . To develop a computer vision application for automated recycling, we leveraged Unity Perception iv , a toolkit for generating large-scale datasets for perceptioniv github.com/Unity-Technologies/com.unity.perception Computer vision pipeline for an automated recycling application (a), which contains multiple ML models, user input, and image data from various sources.…”

Section: Computer Vision With Real and Synthetic Datamentioning

confidence: 99%

Technology Readiness Levels for Machine Learning Systems

Lavin¹,

Gilligan-Lee²,

Visnjic³

et al. 2021

Preprint

View full text Add to dashboard Cite

The development and deployment of machine learning (ML) systems can be executed easily with modern tools, but the process is typically rushed and means-to-an-end. The lack of diligence can lead to technical debt, scope creep and misaligned objectives, model misuse and failures, and expensive consequences. Engineering systems, on the other hand, follow well-defined processes and testing standards to streamline development for high-quality, reliable results. The extreme is spacecraft systems, where mission critical measures and robustness are ingrained in the development process. Drawing on experience in both spacecraft engineering and ML (from research through product across domain areas), we have developed a proven systems engineering approach for machine learning development and deployment. Our Machine Learning Technology Readiness Levels (MLTRL) framework defines a principled process to ensure robust, reliable, and responsible systems while being streamlined for ML workflows, including key distinctions from traditional software engineering. Even more, MLTRL defines a lingua franca for people across teams and organizations to work collaboratively on artificial intelligence and machine learning technologies. Here we describe the framework and elucidate it with several real world use-cases of developing ML methods from basic research through productization and deployment, in areas such as medical diagnostics, consumer computer vision, satellite imagery, and particle physics.

show abstract

Section: Computer Vision With Real and Synthetic Datamentioning

confidence: 99%

Technology Readiness Levels for Machine Learning Systems

Lavin¹,

Gilligan-Lee²,

Visnjic³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…This method requires a large number of unlabled examples from the target class. [59,42,62] find that simulated data improves detection performance, and the degree of realism and variability of simulation affects the amount of improvement. They consider only small sets of non-deformable man-made objects.…”

Section: Data Augmentation Via Style Transfer Generation and Simulamentioning

confidence: 99%

Synthetic Examples Improve Generalization for Rare Classes

Beery

Liu

Morris

et al. 2020

2020 IEEE Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

The ability to detect and classify rare occurrences in images has important applications-for example, counting rare and endangered species when studying biodiversity, or detecting infrequent traffic scenarios that pose a danger to self-driving cars. Few-shot learning is an open problem: current computer vision systems struggle to categorize objects they have seen only rarely during training, and collecting a sufficient number of training examples of rare events is often challenging and expensive, and sometimes outright impossible. We explore in depth an approach to this problem: complementing the few available training images with ad-hoc simulated data. Our testbed is animal species classification, which has a real-world long-tailed distribution. We analyze the effect of different axes of variation in simulation, such as pose, lighting, model, and simulation method, and we prescribe best practices for efficiently incorporating simulated data for real-world performance gain. Our experiments reveal that synthetic data can considerably reduce error rates for classes that are rare, that as the amount of simulated data is increased, accuracy on the target class improves, and that high variation of simulated data provides maximum performance gain.

show abstract

“…As a result, one of the main limitations of modern deep learning-based approaches to quality control are the vast amounts of training data required to develop such solutions, which require considerable human effort, are costly, time consuming and error-prone. In this light, the usage of synthetic data is emerging as an attractive solution to decrease the burden of data collection and annotation [6].…”

Section: A Machine Learning In Quality Controlmentioning

confidence: 99%

Simulation-Based Data Augmentation for the Quality Inspection of Structural Adhesive With Deep Learning

Peres

Guedes²,

Miranda³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

The advent of Industry 4.0 has shown the tremendous transformative potential of combining artificial intelligence, cyber-physical systems and Internet of Things concepts in industrial settings. Despite this, data availability is still a major roadblock for the successful adoption of data-driven solutions, particularly concerning deep learning approaches in manufacturing. Specifically in the quality control domain, annotated defect data can often be costly, time-consuming and inefficient to obtain, potentially compromising the viability of deep learning approaches due to data scarcity. In this context, we propose a novel method for generating annotated synthetic training data for automated quality inspections of structural adhesive applications, validated in an industrial cell for automotive parts. Our approach greatly reduces the cost of training deep learning models for this task, while simultaneously improving their performance in a scarce manufacturing data context with imbalanced training sets by 3.1%

show abstract

An Annotation Saved is an Annotation Earned: Using Fully Synthetic Training for Object Detection

Cited by 97 publications

References 27 publications

Technology Readiness Levels for Machine Learning Systems

Technology Readiness Levels for Machine Learning Systems

Synthetic Examples Improve Generalization for Rare Classes

Simulation-Based Data Augmentation for the Quality Inspection of Structural Adhesive With Deep Learning

Contact Info

Product

Resources

About