Deep learning based models are relatively large, and it is hard to deploy such models on resource-limited devices such as mobile phones and embedded devices. One possible solution is knowledge distillation whereby a smaller model (student model) is trained by utilizing the information from a larger model (teacher model). In this paper, we present an outlook of knowledge distillation techniques applied to deep learning models. To compare the performances of different techniques, we propose a new metric called distillation metric which compares different knowledge distillation solutions based on models' sizes and accuracy scores. Based on the survey, some interesting conclusions are drawn and presented in this paper including the current challenges and possible research directions.

Traffic signs are essential for the safe and efficient movement of vehicles through the transportation network. Poor sign visibility can lead to accidents. One of the key properties used to measure the visibility of a traffic sign is retro-reflection, which indicates how much light a traffic sign reflects back to the driver. The retro-reflection of the traffic sign degrades over time until it reaches a point where the traffic sign has to be changed or repaired. Several studies have explored the idea of modeling the sign degradation level to help the authorities in effective scheduling of sign maintenance. However, previous studies utilized simpler models and proposed multiple models for different combinations of the sheeting type and color used for the traffic sign. In this study, we present a neural network based deep learning model for traffic sign retro-reflectivity prediction. Data utilized in this study was collected using a handheld retro-reflectometer GR3 from field surveys of traffic signs. Sign retro-reflective measurements (i.e., the $R_A$ values) were taken for different sign sheeting brands, grades, colors, orientation angles, observation angles, and aging periods. Feature-based sensitivity analysis was conducted to identify variables’ relative importance in determining retro-reflectivity. Results show that the sheeting color and observation angle were the most significant variables, whereas sign orientation was the least important. Considering all the features, $R_A$ prediction results obtained from one-hot encoding outperformed other models reported in the literature. The findings of this study demonstrate the feasibility and robustness of the proposed neural network based deep learning model in predicting the sign retro-reflectivity.

Due to the scalability of the stateless and compact JSON Web Tokens (JWT), they are increasingly used in securing modern applications to support single-sign-on context and establish trust in microservices or connected-devices architectures. However, currently JWT relies on traditional encryption algorithms that are based on integer factorization or discrete logarithm problems; this exposes JWT signatures to vulnerability when quantum computers become available. Recently, NIST has announced new standards which are quantum safe. In this paper, we investigate the usage of two quantum-resistant algorithms, which are NIST candidates of round 2 standardization process in 2019, to create and verify JWT signatures. Namely, we consider lattice-based schemes (DILITHIUM and qTESLA) and compare their performance against RSA for digital signature for different security levels defined by NIST. The results show that lattice-based digital signature schemes have better performance than RSA in terms of request per seconds and average response time. DILITHIUM has demonstrated the best performance while also having smaller performance loss when scaling the security level.

Unlike sub-metering, which requires individual appliances to be equipped with their own meters, non-intrusive load monitoring (NILM) use algorithms to discover appliance individual consumption from the aggregated overall energy reading. Approaches that uses low frequency sampled data are more applicable in a real world smart meters that has typical sampling capability of <= 1Hz. In this paper, a systematic literature review on deep-learning-based approaches for NILM problem is conducted, aiming to analyse the four key aspects pertaining to deep learning adoption. This includes deep learning model adoption, features selection that are used to train the model, used data set and model accuracy. In our study, analyses the performance of four different deep learning approaches, namely, denoising autoencoder (DAE), recurrent long short-term memory (LSTM) , Recurrent gate recurrent unit (GRU), and sequence to point. Our experiments will be conducted using the two data sets, namely, REDD and UK-DALE. According to our analysis, the sequence to point model has achieved the best results with an average mean absolute error (MAE) of 14:98 watt when compared to other counterpart algorithms.