Explainable Sequence-to-Sequence GRU Neural Network for Pollution Forecasting
The goal of pollution forecasting models is to allow the prediction and control of the air quality. While such deep learning models were deemed for a long time as black boxes, recent advances in eXplainable AI (XAI) allow to look through the...
Optimizing Explanations by Network Canonization and Hyperparameter Search
Rule-based and modified backpropagation XAI methods struggle with innovative layer building blocks and implementation-invariance issues.
In this work we propose canonizations for popular deep neural network architectures and...
Experimental Demonstration of Optical Modulation Format Identification Using SOI-based Photonic Reservoir
We experimentally show modulation format identification in the optical domain using Silicon-on-Insulator-based Photonic-Integrated-Circuit (PIC) reservoir. Identification of 32 GBd single-polarization signals of 4QAM, 16QAM, 32QAM and 64QAM is...
Comparison of Polarization Diversity Configurations of SOI Strip Waveguide-Based Dual-Polarization Wavelength Conversion for S-Band Transmission
Using wavelength conversion of our fabricated SOI strip waveguide, we compared experimentally the polarization-insensitive configuration toward S-band real-time transmission. It is found that parallel configuration is 3dB superior in in-out...
Increasing the power and spectral efficiencies of an OFDM-based VLC system through multi-objective optimization
In order to minimize power usage and maximize spectral efficiency in visible light communication (VLC), we use a multi-objective optimization algorithm and compare DC-biased optical OFDM (DCO-OFDM) with constant envelope OFDM (CE-OFDM)...
Fooling State-of-the-Art Deepfake Detection with High-Quality Deepfakes
Due to the rising threat of deepfakes to security and privacy, it is most important to develop robust and reliable detectors. In this paper, we examine the need for high-quality samples in the training datasets of such detectors. Accordingly, we...
Assessing the Value of Multimodal Interfaces: A Study on Human–Machine Interaction in Weld Inspection Workstations
Multimodal user interfaces promise natural and intuitive human–machine interactions. However, is the extra effort for the development of a complex multisensor system justified, or can users also be satisfied with only one input modality? This...
Shortcomings of Top-Down Randomization-Based Sanity Checks for Evaluations of Deep Neural Network Explanations
While the evaluation of explanations is an important step towards trustworthy models, it needs to be done carefully, and the employed metrics need to be well-understood. Specifically model randomization testing is often overestimated and regarded...
Revealing Hidden Context Bias in Segmentation and Object Detection through Concept-specific Explanations
Applying traditional post-hoc attribution methods to segmentation or object detection predictors offers only limited insights, as the obtained feature attribution maps at input level typically resemble the models' predicted segmentation mask or...
Reveal to Revise: An Explainable AI Life Cycle for Iterative Bias Correction of Deep Models
State-of-the-art machine learning models often learn spurious correlations embedded in the training data. This poses risks when deploying these models for high-stake decision-making, such as in medical applications like skin cancer detection. To...