Neural networks have enabled state-of-the-art approaches to achieve incredible results on computer vision tasks such as object detection. However, such success greatly relies on costly computation resources, which hinders people with cheap devices from appreciating the advanced technology. In this paper, we propose Cross Stage Partial Network (CSPNet) to mitigate the problem that previous works require heavy inference computations from the network architecture perspective. We attribute the problem to the duplicate gradient information within network optimization. The proposed networks respect the variability of the gradients by integrating feature maps from the beginning and the end of a network stage, which, in our experiments, reduces computations by 20% with equivalent or even superior accuracy on the ImageNet dataset, and significantly outperforms state-of-the-art approaches in terms of AP50 on the MS COCO object detection dataset. The CSPNet is easy to implement and general enough to cope with architectures based on ResNet, ResNeXt, and DenseNet.
Cyber-Physical Systems (CPS) are integrations of computation and physical processes. Embedded computers and networks monitor and control the physical processes, usually with feedback loops where physical processes affect computations and vice versa. The economic and societal potential of such systems is vastly greater than what has been realized, and major investments are being made worldwide to develop the technology. There are considerable challenges, particularly because the physical components of such systems introduce safety and reliability requirements qualitatively different from those in general- purpose computing. Moreover, physical components are qualitatively different from object-oriented software components. Standard abstractions based on method calls and threads do not work. This paper examines the challenges in designing such systems, and in particular raises the question of whether today's computing and networking technologies provide an adequate foundation for CPS. It concludes that it will not be sufficient to improve design processes, raise the level of abstraction, or verify (formally or otherwise) designs that are built on today's abstractions. To realize the full potential of CPS, we will have to rebuild computing and networking abstractions. These abstractions will have to embrace physical dynamics and computation in a unified way.
Per- and polyfluoroalkyl substances (PFAS) have drawn public concern recently due to their toxic properties and persistence in the environment, making it urgent to eliminate PFAS from contaminated water. Electrochemical oxidation (EO) has shown great promise for the destructive treatment of PFAS with direct electron transfer and hydroxyl radical (⋅OH)-mediated indirect reactions. One of the most popular electrodes is Magnéli phase titanium suboxides. However, the degradation mechanisms of PFAS are still unsure and are under investigation now. The main methodology is the first-principal density functional theory (DFT) computation, which is recently used to explore the degradation mechanisms and interpret by-product formation during PFAS mineralization. From the literature review, the main applications of DFT computation for studying PFAS degradation mechanisms by EO include bond dissociation energy, absorption energy, activation energy, and overpotential η for oxygen evolution reactions. The main degradation mechanisms and pathways of PFAS in the EO process include mass transfer, direct electron transfer, decarboxylation, peroxyl radical generation, hydroxylation, intramolecular rearrangement, and hydrolysis. In the recent 4 years, 11 papers performed DFT computation to explore the possible PFAS degradation mechanisms and pathways in the EO process. This paper’s objectives are to: 1) summarize the main degradation mechanisms of PFAS degradation in EO; 2) review the application of DFT computation for studying PFAS degradation mechanisms during EO; process; 3) review the possible degradation pathways of perfluorooctane sulfonoic acid (PFOS) and per-fluorooctanoic acid (PFOA) during EO process.
Tobias Heimann, Lara-Sophie Wähling, Tomke Honkomp
et al.
Bioenergy with carbon capture and storage (BECCS) is a crucial element in most modelling studies on emission pathways of the Intergovernmental Panel on Climate Change to limit global warming. BECCS can substitute fossil fuels in energy production and reduce CO _2 emissions, while using biomass for energy production can have feedback effects on land use, agricultural and forest products markets, as well as biodiversity and water resources. To assess the former pros and cons of BECCS deployment, interdisciplinary model approaches require detailed estimates of technological information related to BECCS production technologies. Current estimates of the cost structure and capture potential of BECCS vary widely due to the absence of large-scale production. To obtain more precise estimates, a global online expert survey ( N = 32) was conducted including questions on the regional development potential and biomass use of BECCS, as well as the future operating costs, capture potential, and scalability in different application sectors. In general, the experts consider the implementation of BECCS in Europe and North America to be very promising and regard BECCS application in the liquid biofuel industry and thermal power generation as very likely. The results show significant differences depending on whether the experts work in the Global North or the Global South. Thus, the findings underline the importance of including experts from the Global South in discussions on carbon dioxide removal methods. Regarding technical estimates, the operating costs of BECCS in thermal power generation were estimated in the range of 100–200 USD/tCO _2 , while the CO _2 capture potential was estimated to be 50–200 MtCO _2 yr ^−1 by 2030, with cost-efficiency gains of 20% by 2050 due to technological progress. Whereas the individuals’ experts provided more precise estimates, the overall distribution of estimates reflected the wide range of estimates found in the literature. For the cost shares within BECCS, it was difficult to obtain consistent estimates. However, due to very few current alternative estimates, the results are an important step for modelling the production sector of BECCS in interdisciplinary models that analyse cross-dimensional trade-offs and long-term sustainability.
Given the recent success of metasurfaces in free-space applications, these concepts can be leveraged to an even larger extent in on-chip waveguide systems. The in-plane diffractive metasurfaces enable the manipulation of guiding waves in the multimode regime with greater parallelism than conventional single-mode or few-mode waveguides, leading to exciting opportunities in signal processing and optical computing systems. Beam focusing is one of the basic functionalities of wavefront shaping, which can be implemented using phase gradient metalenses consisting of arrays of meta-atoms. The meta-atoms are mainly realized by etched trenches with varying lengths, which are assembled into a one-dimensional transmit array with a specific phase response. However, this kind of periodic arrayed structure has significantly limited design freedom compared to its free-space counterparts. Here, we propose a digital metalens that consists of a seamless array of pixelated unit cells, which are engineered via inverse design. In contrast to conventional focusing metalenses based on transmit arrays, highly functional digital metalenses have been demonstrated: (1) achromatic focusing lens; (2) extended depth of focus (EDOF) lens; (3) Airy beam lens. These devices were fabricated on a silicon photonic platform and characterized in near-infrared. The intersection of digital structures and algorithm-driven optimizations offers greater versatility for on-chip wavefront shaping.
Ching-Wei Wang, Hikam Muzakky, Yu-Ching Lee
et al.
Thyroid cancer is the most common endocrine cancer. Papillary thyroid cancer (PTC) is the most prevalent form of malignancy among all thyroid cancers arising from follicular cells. Fine needle aspiration cytology (FNAC) is a non-invasive method regarded as the most cost-effective and accurate diagnostic method of choice in diagnosing PTC. Identification of BRAF (V600E) mutation in thyroid neoplasia may be beneficial because it is specific for malignancy, implies a worse prognosis, and is the target for selective BRAF inhibitors. To the authors’ best knowledge, this is the first automated precision oncology framework effectively predict BRAF (V600E) immunostaining result in thyroidectomy specimen directly from Papanicolaou-stained thyroid fine-needle aspiration cytology and ThinPrep cytological slides, which is helpful for novel targeted therapies and prognosis prediction. The proposed deep learning (DL) framework is evaluated on a dataset of 118 whole slide images. The results show that the proposed DL-based technique achieves an accuracy of 87%, a precision of 94%, a sensitivity of 91%, a specificity of 71% and a mean of sensitivity and specificity at 81% and outperformed three state-of-the-art deep learning approaches. This study demonstrates the feasibility of DL-based prediction of critical molecular features in cytological slides, which not only aid in accurate diagnosis but also provide useful information in guiding clinical decision-making in patients with thyroid cancer. With the accumulation of data and the continuous advancement of technology, the performance of DL systems is expected to be improved in the near future. Therefore, we expect that DL can provide a cost-effective and time-effective alternative tool for patients in the era of precision oncology.