The emergence of a pandemic affecting the respiratory system can result in a significant demand for face masks. This includes the use of cloth masks by large sections of the public, as can be seen during the current global spread of COVID-19. However, there is limited knowledge available on the performance of various commonly available fabrics used in cloth masks. Importantly, there is a need to evaluate filtration efficiencies as a function of aerosol particulate sizes in the 10 nm to 10 μm range, which is particularly relevant for respiratory virus transmission. We have carried out these studies for several common fabrics including cotton, silk, chiffon, flannel, various synthetics, and their combinations. Although the filtration efficiencies for various fabrics when a single layer was used ranged from 5 to 80% and 5 to 95% for particle sizes of <300 nm and >300 nm, respectively, the efficiencies improved when multiple layers were used and when using a specific combination of different fabrics. Filtration efficiencies of the hybrids (such as cotton−silk, cotton−chiffon, cotton−flannel) was >80% (for particles <300 nm) and >90% (for particles >300 nm). We speculate that the enhanced performance of the hybrids is likely due to the combined effect of mechanical and electrostatic-based filtration. Cotton, the most widely used material for cloth masks performs better at higher weave densities (i.e., thread count) and can make a significant difference in filtration efficiencies. Our studies also imply that gaps (as caused by an improper fit of the mask) can result in over a 60% decrease in the filtration efficiency, implying the need for future cloth mask design studies to take into account issues of "fit" and leakage, while allowing the exhaled air to vent efficiently. Overall, we find that combinations of various commonly available fabrics used in cloth masks can potentially provide significant protection against the transmission of aerosol particles.

Humans and most animals can learn new tasks without forgetting old ones. However, training artificial neural networks (ANNs) on new tasks typically causes them to forget previously learned tasks. This phenomenon is the result of "catastrophic forgetting", in which training an ANN disrupts connection weights that were important for solving previous tasks, degrading task performance. Several recent studies have proposed methods to stabilize connection weights of ANNs that are deemed most important for solving a task, which helps alleviate catastrophic forgetting. Here, drawing inspiration from algorithms that are believed to be implemented in vivo, we propose a complementary method: adding a context-dependent gating signal, such that only sparse, mostly non-overlapping patterns of units are active for any one task. This method is easy to implement, requires little computational overhead, and allows ANNs to maintain high performance across large numbers of sequentially presented tasks, particularly when combined with weight stabilization. We show that this method works for both feedforward and recurrent network architectures, trained using either supervised or reinforcement-based learning. This suggests that employing multiple, complimentary methods, akin to what is believed to occur in the brain, can be a highly effective strategy to support continual learning.The goal of this study was to develop neuroscience-inspired methods to alleviate catastrophic forgetting in ANNs. Two previous studies have proposed one such method: stabilizing connection weights depending on their importance for solving a task [8,9]. This method, inspired by neuroscience research demonstrating that stabilization of dendritic spines is associated with task learning and retention [6,7], has shown promising results when trained and tested on sequences of ≤ 10 tasks. However, it is uncertain how well these methods perform when trained on much larger number of sequential tasks.We first tested whether these methods can alleviate catastrophic forgetting by measuring performance on 100 sequentially presented digit classification tasks. Specifically, we tested a fully connected feedforward network with two hidden layers (2000 units each, Figure 1A) on the permuted MNIST digit classification task [13]. This involved training the network on the MNIST task for 20 epochs, permuting the 784 pixels in all images with the same permutation, and then training the network on this new set of images. This test is a canonical example of an "input reformatting" problem, in which the input and output semantics (pixel intensities and digit identity, respectively) are identical across all tasks, but the input format (the spatial location of each pixel) changes between tasks [13].We sequentially trained the base ANN on 100 permutations of the image set. Without any synaptic stabilization, this network can classify digits with an accuracy of 98.5% for any single permutation, but mean classification accuracy falls to 52.5% after the network is trained on 10 permutation...

No abstract

No abstract

Isolated solid-state atomic defects with telecom optical transitions are ideal quantum photon emitters and spin qubits for applications in long-distance quantum communication networks. Prototypical telecom defects, such as erbium, suffer from poor photon emission rates, requiring photonic enhancement using resonant optical cavities. Moreover, many of the traditional hosts for erbium ions are not amenable to direct incorporation with existing integrated photonics platforms, limiting scalable fabrication of qubit-based devices. Here, we present a scalable approach toward CMOS-compatible telecom qubits by using erbium-doped titanium dioxide thin films grown atop silicon-oninsulator substrates. From this heterostructure, we have fabricated onedimensional photonic crystal cavities demonstrating quality factors in excess of 5 × 10 4 and corresponding Purcell-enhanced optical emission rates of the erbium ensembles in excess of 200. This easily fabricated materials platform represents an important step toward realizing telecom quantum memories in a scalable qubit architecture compatible with mature silicon technologies.