SYNTA: A novel approach for deep learning-based image analysis in muscle histopathology using photo-realistic synthetic data.
Leonid MillOliver AustJochen A. AckermannPhilipp BurgerMonica PascualKatrin Palumbo-ZerrGerhard KrönkeStefan UderhardtGeorg SchettChristoph S. ClemenRolf SchröderChristian HoltzhausenSamir JabariAndreas K. MaierAnika GrüneboomPublished in: CoRR (2022)
Keyphrases
- synthetic data
- deep learning
- photorealistic
- image analysis
- virtual reality
- computer graphics
- hand held
- d scene
- unsupervised learning
- highly accurate
- image space
- range images
- facial expression synthesis
- complex scenes
- machine learning
- real world
- pattern recognition
- image acquisition
- vision system
- computer vision
- data sets
- image segmentation
- mental models
- three dimensional
- real image data
- data mining
- point cloud
- medical imaging
- image processing
- high quality
- supervised learning
- weakly supervised
- d objects
- object detection