Rotation Invariant Networks for Image Classification for HPC and Embedded Systems

Abstract

Convolutional Neural Network (CNNs) models’ size reduction has recently gained interest due to several advantages: energy cost reduction, embedded devices, and multi-core interfaces. One possible way to achieve model reduction is the usage of Rotation-invariant Convolutional Neural Networks because of the possibility of avoiding data augmentation techniques. In this work, we present the next step to obtain a general solution to endowing CNN architectures with the capability of classifying rotated objects and predicting the rotation angle without data-augmentation techniques. The principle consists of the concatenation of a representation mapping transforming rotation to translation and a shared weights predictor. This solution has the advantage of admitting different combinations of various basic, existing blocks. We present results obtained using a Gabor-filter bank and a ResNet feature backbone compared to previous other solutions. We also present the possibility to select between parallelizing the network in several threads for energy-aware High Performance Computing (HPC) applications or reducing the memory footprint for embedded systems. We obtain a competitive error rate on classifying rotated MNIST and outperform existing state-of-the-art results on CIFAR-10 when trained on up-right examples and validated on random orientations.

Motion blur invariant for estimating motion parameters of medical ultrasound images

“…This results in a trade-off between improving image quality and minimizing motion blurring. Invariant features play an important role in pattern recognition and shape detection as image characteristics may stand consistent under particular transformations such as scaling, translation and rotation 17 , 18 . Along with some research targeting to estimate the PSF, noise and speckle parameters (see 19 for example), there are several successful approaches using the theory of image invariant analysis for ultrasound image segmentation and classification (see 20 and 21 for a survey).…”