18384549. METHOD AND ELECTRONIC DEVICE FOR TRAINING NEURAL NETWORK MODEL BY AUGMENTING IMAGE REPRESENTING OBJECT CAPTURED BY MULTIPLE CAMERAS simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)
METHOD AND ELECTRONIC DEVICE FOR TRAINING NEURAL NETWORK MODEL BY AUGMENTING IMAGE REPRESENTING OBJECT CAPTURED BY MULTIPLE CAMERAS
Organization Name
Inventor(s)
METHOD AND ELECTRONIC DEVICE FOR TRAINING NEURAL NETWORK MODEL BY AUGMENTING IMAGE REPRESENTING OBJECT CAPTURED BY MULTIPLE CAMERAS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18384549 titled 'METHOD AND ELECTRONIC DEVICE FOR TRAINING NEURAL NETWORK MODEL BY AUGMENTING IMAGE REPRESENTING OBJECT CAPTURED BY MULTIPLE CAMERAS
Simplified Explanation
The computer-implemented method described in the abstract involves training a neural network model by augmenting images representing objects. Here is a simplified explanation of the abstract:
- Obtaining object recognition results from images captured by different cameras.
- Converting the object recognition results between different camera coordinate systems.
- Generating training data based on the converted object recognition results.
- Training a neural network model using the generated training data.
- Potential Applications
This technology could be applied in various fields such as autonomous driving, surveillance systems, and robotics for object recognition and training neural networks.
- Problems Solved
This technology addresses the challenge of training neural network models with diverse viewpoints of objects captured by different cameras, improving the model's accuracy and robustness.
- Benefits
The benefits of this technology include enhanced object recognition capabilities, improved training data quality, and increased performance of neural network models in real-world scenarios.
- Potential Commercial Applications
Potential commercial applications of this technology include developing advanced security systems, optimizing industrial processes, and enhancing medical imaging technologies.
- Possible Prior Art
One possible prior art could be the use of data augmentation techniques in computer vision to improve the performance of neural networks trained on images captured from different viewpoints.
- Unanswered Questions:
1. How does the method handle variations in lighting conditions between images captured by different cameras? 2. Are there any limitations in the size or type of objects that can be effectively recognized and trained using this method?
Original Abstract Submitted
Provided is a computer-implemented method of training a neural network model by augmenting images representing objects. The method includes: obtaining a first object recognition result predicted by a first neural network model using, as an input, a first image captured by a first camera capturing, from a first viewpoint, a space including at least one object; converting the obtained first object recognition result, based on a conversion relationship between a first camera coordinate system corresponding to the first camera and a second camera coordinate system corresponding to a second camera capturing, from a second viewpoint, the space; generating, based on the first object recognition result converted with respect to the second viewpoint, training data by performing labeling on a second image that corresponds to the first image, the second image being captured by the second camera; and training a second neural network model by using the generated training data.
