Data enhancement including background modification for robust prediction using neural networks

In various examples, a background of an object may be modified to generate a training image. A segmentation mask may be generated and used to generate an object image including image data representing an object. Object images may be integrated into different backgrounds and used to train data enhancements in neural networks. Data enhancement may also be performed using hue adjustment (e.g., of an object image) and/or rendering of three-dimensional captured data corresponding to an object from a selected view. The inference score may be analyzed to select a background of the image to be included in the training data set. A background may be selected and training images may be iteratively added to a training dataset during training (e.g., between periods). In addition, early or late fusion using object mask data may be employed to improve reasoning performed by neural networks trained using object mask data. 在各种示例中，可以修改对象的背景以生成训练图像。分割掩码可被生成并用于生成包括表示对象的图像数据的对象图像。对象图像可以集成到不同的背景中，并用于训练神经网络中的数据增强。还可以使用色调调整(例如，对象图像的