Search

Precision viticulture is an area that is very dependent on methods that allow for a sustainable assessment of grape maturity and, in this work, we apply two state-of-the-art (SOTA) convolution-based networks, namely InceptionTime and OmniScale 1D-CNN, to hyperspectral images of wine grape berries to estimate sugar content. Since attaining generalization capacity and processing the information in such high-dimensional data are the two biggest challenges to overcome in problems of this nature, we also study the impact of two dimensionality reduction techniques, Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE), on the models' performance. Both models underwent different tests with different vintages and varieties of wine grapes in the training/validation steps, as to form a true test to their generalization capacity. Our results show that both PCA and t-SNE succeed in improving the performance of these deep networks when an adequate number of components is chosen that minimizes the ratio between information loss and removing redundant features: additionally, both techniques significantly reduce computational cost, a very important trait when training deep learning models. Both models showed good generalization ability with very competitive results across different varieties and vintages even despite their significant differences in variability, which is an indicator that a relationship between spectras can be found that is reflected on sugar content values.

Motion or inertial sensors such as gyroscope and accelerometer commonly found in smartwatches and smartphones can measure characteristics such as acceleration and angular velocity of movements in the human body and use them to learn models capable of identifying human activities, that has applicability in various fields such as biometrics, remote patient health monitoring, etc. Recently deep learning-based methods have become popular for human activity recognition because they use representation learning techniques that can automatically generate optimal features from raw input data generated from sensors without any human intervention and can identify hidden patterns in data. This work proposes a novel hybrid deep neural network model, CNN-GRU that combines convolutional and gated recurrent units for human activity recognition. This model was successfully validated on WISDM dataset and produced accuracy that is suggestively better than other state-of-the-art deep neural network models such as Inception Time and DeepConvLSTM created using AutoML.