Visual defect obfuscation based self-supervised anomaly detection.

Due to scarcity of anomaly situations in the early manufacturing stage, an unsupervised anomaly detection (UAD) approach is widely adopted which only uses normal samples for training. This approach is based on the assumption that the trained UAD model will accurately reconstruct normal patterns but struggles with unseen anomalies. To enhance the UAD performance, reconstruction-by-inpainting based methods have recently been investigated, especially on the masking strategy of suspected defective regions. However, there are still issues to overcome: (1) time-consuming inference due to multiple masking, (2) output inconsistency by random masking, and (3) inaccurate reconstruction of normal patterns for large masked areas. Motivated by this, this study proposes a novel reconstruction-by-inpainting method, dubbed Excision And Recovery (EAR), that features single deterministic masking based on the ImageNet pre-trained DINO-ViT and visual obfuscation for hint-providing. Experimental results on the MVTec AD dataset show that deterministic masking by pre-trained attention effectively cuts out suspected defective regions and resolves the aforementioned issues 1 and 2. Also, hint-providing by mosaicing proves to enhance the performance than emptying those regions by binary masking, thereby overcomes issue 3. The proposed approach achieves a high performance without any change of the model structure. Promising results are shown through laboratory tests with public industrial datasets. To suggest EAR be possibly adopted in various industries as a practically deployable solution, future steps include evaluating its applicability in relevant manufacturing environments.

Insight into the latitudinal gradient of biodiversity based on spatial variations in pelagic ciliate communities along the western Arctic Ocean.

The latitudinal dynamics of biodiversity has been the focus of global attention. This study is based on the latitude gradient of biodiversity in the spatial changes of pelagic ciliate communities in the western Arctic Ocean. The gradient pattern of pelagic ciliate communities across four latitudes were investigated from the water surface at 22 sampling station in the northern Bering Sea of the western Arctic Ocean and Chukchi Sea from August 5 to August 24, 2016. Based on multivariate analyses, the results showed that (1) the spatial patterns of pelagic ciliates represented a significant latitudinal gradient along the western Arctic Ocean; (2) the species number and abundance of pelagic ciliate communities declined from 64°N to 80°N; (3) variations in the horizontal distribution of ciliates were significantly correlated with changes in physicochemical variables, especially water temperature and Chl a; Thus it is suggested that the expected latitudinal decline of biodiversity was evident along the western Arctic Ocean.

Atlantic-origin water extension into the Pacific Arctic induced an anomalous biogeochemical event.

The Arctic Ocean is facing dramatic environmental and ecosystem changes. In this context, an international multiship survey project was undertaken in 2020 to obtain current baseline data. During the survey, unusually low dissolved oxygen and acidified water were found in a high-seas fishable area of the western (Pacific-side) Arctic Ocean. Herein, we show that the Beaufort Gyre shrinks to the east of an ocean ridge and forms a front between the water within the gyre and the water from the eastern (Atlantic-side) Arctic. That phenomenon triggers a frontal northward flow along the ocean ridge. This flow likely transports the low oxygen and acidified water toward the high-seas fishable area; similar biogeochemical properties had previously been observed only on the shelf-slope north of the East Siberian Sea.

Factors affecting the subsurface aragonite undersaturation layer in the Pacific Arctic region.

This study evaluated interannual variation in the subsurface aragonite undersaturation zone (ΩAr<1 layer) in the Pacific Arctic Ocean, using data from the 2016-2019 period. The upper boundary (DEPΩ<1UB) of the ΩAr<1 layer generally formed at a depth where the contribution of corrosive Pacific water was approximately 98 %. The intensity of the Beaufort Gyre associated with freshwater accumulation mainly determined interannual variation in DEPΩ<1UB, but the direction of its effect was opposite west and east of ~166°W. The lower boundary (DEPΩ<1LB) of the ΩAr<1 layer was generally found at a depth range where equal contributions of Pacific and Atlantic water were expected. An Atlantic-origin cold saline water intrusion event in 2017 caused by an anomalous atmospheric circulation pattern significantly lifted the DEPΩ<1LB, thus the thickness of the ΩAr<1 layer decreased.

Zero standby power crop water-stress detector leading to the optimization of water usage and yield.

Agricultural sensors are powerful tools to optimize crop productivity while conserving natural resources. Here we report a crop water-stress detector based on a plasmonically-enhanced micromechanical photoswitch capable of detecting water content in leaves that is lower than a predetermined threshold without consuming electrical power when the leaf is healthy. The detection mechanism exploits the energy in a specific narrow-spectral band of solar radiation reflected off leaves that is strongly correlated to the water content in plants. This biosensor relies on a spectrally selective infrared plasmonic absorber and a thermally sensitive micro-cantilever to harvest the reflected solar energy and further produce a digitized wakeup-bit only when the monitored leaf is water-stressed. In particular, we demonstrate that the detector activates a commercial water pump when a soybean plant is water-stressed. The 10-year battery lifetime of the proposed detector pave the way for the development of high-granularity, maintenance-free sensor networks for large-scale smart-farms.

The Interplay of Mycosporine-like Amino Acids between Phytoplankton Groups and Northern Krill (Thysanoessa sp.) in a High-Latitude Fjord (Kongsfjorden, Svalbard).

We investigated pigment and mycosporine-like amino acid (MAA) concentrations of phytoplankton and Northern krill (Thysanoessa sp.) in sub-Arctic Kongsfjorden. Chlorophyll a (Chl-a) concentrations in the surface and middle-layer water were 0.44 µg L-1 (±0.17 µg L-1) and 0.63 µg L-1 (±0.25 µg L-1), respectively. Alloxanthin (Allo, a marker of cryptophytes) was observed at all stations, and its mean values for surface and middle-layer water were 0.09 µg L-1 (±0.05 µg L-1) and 0.05 (±0.02 µg L-1), respectively. The mean MAA-to-Chl-a ratios at the surface (3.31 ± 2.58 µg (µg Chl-a)-1) were significantly higher than those in the middle-layer water (0.88 ± 0.49 µg (µg Chl-a)-1), suggesting that these compounds play an important role in reducing UV photodamage. In gut pigment levels of Northern krill, the most abundant accessory pigment was Allo (2.79 ± 0.33 µg g-1 dry weight; d.w.), as was the accumulation of Chl-a (8.29 ± 1.13 µg g-1 d.w.). The average concentration of MAAs was 1.87 mg g-1 d.w. (±0.88 mg g-1 d.w.) in krill eyes, which was higher than that in all other body parts (0.99 ± 0.41 mg g-1 d.w.), except for the gut. Thysanoessa sp. was found to contain five identified MAAs (shinorine, palythine, porphyra-334, mycosporine-glycine, and M-332) in the krill eye, whereas shinorine and porphyra-334 were only observed in the krill body, not the eyes and gut. These findings suggest that Northern krill accumulate MAAs of various compositions through the diet (mainly cryptophytes) and translocate them among their body parts as an adaptation for photoprotection and physiological demands.

Low-Power Scan Correlation-Aware Scan Cluster Reordering for Wireless Sensor Networks.

Cryptographic circuits generally are used for applications of wireless sensor networks to ensure security and must be tested in a manufacturing process to guarantee their quality. Therefore, a scan architecture is widely used for testing the circuits in the manufacturing test to improve testability. However, during scan testing, test-power consumption becomes more serious as the number of transistors and the complexity of chips increase. Hence, the scan chain reordering method is widely applied in a low-power architecture because of its ability to achieve high power reduction with a simple architecture. However, achieving a significant power reduction without excessive computational time remains challenging. In this paper, a novel scan correlation-aware scan cluster reordering is proposed to solve this problem. The proposed method uses a new scan correlation-aware clustering in order to place highly correlated scan cells adjacent to each other. The experimental results demonstrate that the proposed method achieves a significant power reduction with a relatively fast computational time compared with previous methods. Therefore, by improving the reliability of cryptography circuits in wireless sensor networks (WSNs) through significant test-power reduction, the proposed method can ensure the security and integrity of information in WSNs.

Importance of seasonal sea ice in the western Arctic ocean to the Arctic and global microplastic budgets.

Arctic sea ice entraps microplastics (MP) from seawater and atmosphere and is recognized as sink and transport vector of MPs. However, ice-trapped fraction in the global MP budget, contribution of atmospheric input, and linkage among Arctic basins remain unclear. To assess them, we investigated the number- and mass-based data separated by size and shape geometry for MPs in sea ice, snow, and melt pond water from the western Arctic Ocean (WAO). A significant dependency of MP data on measured cutoff size and geometry was found. For the same size range and geometry, sea ice MPs in WAO ((11.4 ± 9.12) × 103 N m-3 for ≥ 100 µm) were within comparable levels with those in other Arctic basins, but showed closer similarity in polymer and shape compositions between WAO and Arctic Central Basin, indicating the strong linkage of the two basins by the Transpolar Drift. Our budgeting shows that a significant amount of plastic particles ((3.4 ± 2.6) × 1016 N; 280 ± 701 kilotons), which are missed from the global inventory, is trapped in WAO seasonal sea ice, with < 1% snowfall contribution. Our findings highlight that WAO ice zone may play a role as a sink of global MPs as well as a source of Arctic MPs.

N2O dynamics in the western Arctic Ocean during the summer of 2017.

The western Arctic Ocean (WAO) has experienced increased heat transport into the region, sea-ice reduction, and changes to the WAO nitrous oxide (N2O) cycles from greenhouse gases. We investigated WAO N2O dynamics through an intensive and precise N2O survey during the open-water season of summer 2017. The effects of physical processes (i.e., solubility and advection) were dominant in both the surface (0-50 m) and deep layers (200-2200 m) of the northern Chukchi Sea with an under-saturation of N2O. By contrast, both the surface layer (0-50 m) of the southern Chukchi Sea and the intermediate (50-200 m) layer of the northern Chukchi Sea were significantly influenced by biogeochemically derived N2O production (i.e., through nitrification), with N2O over-saturation. During summer 2017, the southern region acted as a source of atmospheric N2O (mean: + 2.3 ± 2.7 µmol N2O m-2 day-1), whereas the northern region acted as a sink (mean - 1.3 ± 1.5 µmol N2O m-2 day-1). If Arctic environmental changes continue to accelerate and consequently drive the productivity of the Arctic Ocean, the WAO may become a N2O "hot spot", and therefore, a key region requiring continued observations to both understand N2O dynamics and possibly predict their future changes.

Tracing riverine dissolved organic carbon and its transport to the halocline layer in the Chukchi Sea (western Arctic Ocean) using humic-like fluorescence fingerprinting.

Dissolved organic carbon (DOC) and the fluorescence properties of dissolved organic matter (FDOM) were investigated using parallel factor analysis (PARAFAC) for seawater samples collected in the Chukchi Sea (65°N-78°N, 170°E-160°W) during summer 2017. River water (friver) and sea-ice meltwater (fsea ice melt) fractions were also derived using oxygen isotopes ratios (δ18O) to examine the influence of sea ice on riverine DOM. The spatial distributions of friver, riverine DOC, and the humic-like fluorescent component (C1) showed an overall south-north gradient, with higher values in the northern Chukchi Sea in summer. Pronounced accumulation of river water and riverine DOM was also observed in the anticyclonic Beaufort Gyre at the eastern stations of the northern Chukchi Sea in association with a long water residence time. Estimated riverine DOC in the surface layer accounted for 27 ± 9% (range: 17-47%) of the total DOC in the southern Chukchi Sea, and 39 ± 6% (range: 32-49%) and 31 ± 4% (range: 25-37%) for the eastern and western stations of the northern Chukchi Sea, respectively. Humic-like C1 showed negative and positive relationships with sea-ice meltwater-corrected salinity (Ssim_corrected) and friver, respectively. However, Arctic river waters with distinct humic-like C1 characteristics were likely mixed in the northern Chukchi Sea. The vertical distributions of riverine DOC, humic-like C1 fluorescence, and friver generally decreased with water depth, reflecting the strong influence of riverine DOM in the surface layer. Although riverine DOM and friver were dominant in the upper 50 m of the water column, they were also pronounced in the upper halocline (50-200 m), in which fsea ice melt dropped below zero. Our results indicated the existence of brine rejected from growing sea ice, and that sea-ice formation was a key factor for the transport of riverine DOM to the upper halocline layer in the northern Chukchi Sea.

Scan-Chain-Fault Diagnosis Using Regressions in Cryptographic Chips for Wireless Sensor Networks.

Scan structures, which are widely used in cryptographic circuits for wireless sensor networks applications, are essential for testing very-large-scale integration (VLSI) circuits. Faults in cryptographic circuits can be effectively screened out by improving testability and test coverage using a scan structure. Additionally, scan testing contributes to yield improvement by identifying fault locations. However, faults in circuits cannot be tested when a fault occurs in the scan structure. Moreover, various defects occurring early in the manufacturing process are expressed as faults of scan chains. Therefore, scan-chain diagnosis is crucial. However, it is difficult to obtain a sufficiently high diagnosis resolution and accuracy through the conventional scan-chain diagnosis. Therefore, this article proposes a novel scan-chain diagnosis method using regression and fan-in and fan-out filters that require shorter training and diagnosis times than existing scan-chain diagnoses do. The fan-in and fan-out filters, generated using a circuit logic structure, can highlight important features and remove unnecessary features from raw failure vectors, thereby converting the raw failure vectors to fan-in and fan-out vectors without compromising the diagnosis accuracy. Experimental results confirm that the proposed scan-chain-diagnosis method can efficiently provide higher resolutions and accuracies with shorter training and diagnosis times.

Use of a broad ß-diversity measure of pelagic ciliate communities for assessing vertical heterogeneity of water columns in the Pacific Arctic Ocean.

Multivariate dispersion has proven to be a broad ß-diversity measure that shows the heterogeneity of environmental conditions. The dispersion patterns of pelagic ciliate communities were investigated at eight water depths in the northern Bering Sea of the western Arctic Ocean and Chukchi Sea. Multivariate analysis indicated that (1) pelagic ciliates showed significant variability in multivariate dispersion on a vertical scale, (2) dispersion patterns were shaped by both the species composition and individual abundance, (3) vertical variation in species occurrence was significantly related to nutrients and chlorophyll a, and (4) the dispersion measures at both species occurrence and species abundance resolutions were significantly negatively related to salinity and dissolved oxygen. This suggests that multivariate dispersion measures driven by both species composition and the individual abundance of pelagic ciliates may be a useful indicator of environmental heterogeneity in marine ecosystems.

Contrasting Community Composition of Active Microbial Eukaryotes in Melt Ponds and Sea Water of the Arctic Ocean Revealed by High Throughput Sequencing.

Melt ponds (MPs), form as the result of thawing of snow and sea ice in the summer, have lower albedo than the sea ice and are thus partly responsible for the polar amplification of global warming. Knowing the community composition of MP organisms is key to understanding their roles in the biogeochemical cycles of nutrients and elements. However, the community composition of MP microbial eukaryotes has rarely been studied. In the present study, we assessed the microbial eukaryote biodiversity, community composition, and assembly processes in MPs and surface sea water (SW) using high throughput sequencing of 18S rRNA of size-fractionated samples. Alpha diversity estimates were lower in the MPs than SW across all size fractions. The community composition of MPs was significantly different from that of SW. The MP communities were dominated by members from Chrysophyceae, the ciliate classes Litostomatea and Spirotrichea, and the cercozoan groups Filosa-Thecofilosea. One open MP community was similar to SW communities, which was probably due to the advanced stage of development of the MP enabling the exchange of species between it and adjacent SW. High portions of shared species between MPs and SW may indicate the vigorous exchange of species between these two major types of environments in the Arctic Ocean. SW microbial eukaryote communities are mainly controlled by dispersal limitation whereas those of MP are mainly controlled by ecological drift.

Latitudinal Distributions and Controls of Bacterial Community Composition during the Summer of 2017 in Western Arctic Surface Waters (from the Bering Strait to the Chukchi Borderland).

The western Arctic Ocean is experiencing some of the most rapid environmental changes in the Arctic. However, little is known about the microbial community response to these changes. Employing observations from the summer of 2017, this study investigated latitudinal variations in bacterial community composition in surface waters between the Bering Strait and Chukchi Borderland and the factors driving the changes. Results indicate three distinctive communities. Southern Chukchi bacterial communities are associated with nutrient rich conditions, including genera such as Sulfitobacter, whereas the northern Chukchi bacterial community is dominated by SAR clades, Flavobacterium, Paraglaciecola, and Polaribacter genera associated with low nutrients and sea ice conditions. The frontal region, located on the boundary between the southern and northern Chukchi, is a transition zone with intermediate physical and biogeochemical properties; however, bacterial communities differed markedly from those found to the north and south. In the transition zone, Sphingomonas, with as yet undetermined ecological characteristics, are relatively abundant. Latitudinal distributions in bacterial community composition are mainly attributed to physical and biogeochemical characteristics, suggesting that these communities are susceptible to Arctic environmental changes. These findings provide a foundation to improve understanding of bacterial community variations in response to a rapidly changing Arctic Ocean.

Potential Implications of Changing Photosynthetic End-Products of Phytoplankton Caused by Sea Ice Conditions in the Northern Chukchi Sea.

The recent dramatic decline in sea ice conditions in the Arctic Ocean has led to the ecophysiological changes in the phytoplankton community. Little is currently known about how the physiological status of phytoplankton has changed under rapidly changing environmental conditions in the Arctic Ocean. Using the 13C isotope tracer technique, the carbon allocation of phytoplankton into different photosynthetic end-products was determined in the northern Chukchi Sea on the basis of two Arctic expeditions conducted in 2011 and 2012 to identify the physiological status of phytoplankton. Lipids were the predominant photosynthetic biochemical fraction (42.5%) in 2011, whereas carbon allocation to proteins was most dominant under ice-free conditions in 2012 (47.7%). Based on a comparison of the photosynthetic carbon allocation of phytoplankton according to sea ice conditions, we found that photosynthetic carbon allocation to different macromolecular pools was significantly different depending on the sea ice conditions and that the light conditions caused by different sea ice conditions could be an important reason for the differences in carbon allocation to photosynthetic end-products. Different dominant phytoplankton groups related to size classes also could cause changes in the photosynthetic carbon allocation of phytoplankton related mainly to the lipid synthesis. Our results showed that the physiological status of Arctic phytoplankton could be changed by producing different photosynthetic end-products under current environmental changes. This change in photosynthetic end-products of phytoplankton as a basic food source could be further linked to higher trophic levels in regards to their nutritional and energetic aspects, which could have potential consequences for Arctic marine ecosystems.

A low-cost concurrent TSV test architecture with lossless test output compression scheme.

As the traditional IC design migrates to three-dimensional integrated circuits (3D-ICs) design, new challenges need to be considered carefully to solve its reliability and yield issues. 3D-ICs using through-silicon-vias (TSVs) can have latent defects such as resistive open and bridge defects, which are caused by the thermal stress during the fabrication process. These latent defects lead to the deterioration of the electrical performance of TSVs caused by an undesired increase in the resistance-capacitance (RC) delay. For this reason, various post-bond test methodologies have been studied to improve the reliability of 3D-ICs. Cost reduction in these TSV test architectures is also currently being studied by decreasing various factors such as hardware overhead, test time, and the peak current consumption. Usually, a single test-clock-period is required to determine whether the test result contains the defective TSV. When the test result of any TSVs fails, we use another single test-clock-period to classify its defect type. In this paper, we propose a new TSV test architecture to transfer the combined test output of the test result and the specific defect type to the pad during the single test-clock-period. Our proposed test architecture also provides a reliable block-based concurrent testing to optimize the test time by dividing the die into concurrent blocks. The experimental results showed that our proposed test architecture could reduce the test time and the hardware overhead substantially by ensuring that the reasonable peak power consumption for mass production was reasonable without the test quality being adversely affected.

A debug scheme to improve the error identification in post-silicon validation.

While developing semiconductors, post-silicon validation is an important step to identify the errors that are not detected during the pre-silicon verification and manufacturing testing phases. When the design complexity increases, the required debug time also increases because additional debug data are required to identify the errors. In this study, we present a debug scheme that improves the error identification capability. The proposed debug approach concurrently generates three types of signatures using hierarchical multiple-input signature registers (MISRs). The error-suspect debug cycles are determined by analyzing the debug cycles that are commonly contained in the erroneous signatures of the three MISRs. To reduce the amount of debug data, we compare the high-level MISR signatures in real time with the golden signatures; further, we handle the remaining two MISRs based on the tag bits that are obtained from the results of the high-level MISR. The experimental results prove that the proposed debug structure can significantly improve the error identification capability using less debug data than that used in previous debug structure.

Synthesis of mycosporine-like amino acids by a size-fractionated marine phytoplankton community of the arctic beaufort sea.

During the RV-ARAON cruise, a comparative study on the biosynthesis of mycosporine-like amino acids (MAAs) was conducted for the size-fractionated phytoplankton of the Beaufort Sea (Arctic). The MAAs contents in the micro-phytoplankton community (>20 µm size) is considerably higher than that observed in the nano- (20-2 µm size) and pico-phytoplankton (<2 µm size) communities. The micro-phytoplankton of the Mackenzie Shelf had a relatively higher Chlorophyll a (Chl a) concentration. Considering the total phytoplankton community, the MAAs concentration as well as net production of individual MAAs (such as shinorine and palythine) were higher at the Mackenzie Shelf rather than at the sites located beyond the Beaufort Sea; precisely, the highest net production rates of shinorine and palythine were 0.211 (±0.02) ng C L-1 d-1 and 0.136 (±0.001) ng C L-1 d-1 respectively (No other MAAs were detected). The micro-phytoplankton used around 0.5% of the total carbon uptake for the synthesis of MAAs. Compared to the smaller phytoplankton community, the micro-phytoplankton utilized more of their energy for the biosynthesis of MAAs; on the other hand, nano- and pico-phytoplankton focused on cellular activity and had poor biosynthesis of MAAs. This clearly indicates the phytoplankton size-dependent variation in the biosynthesis of MAA in the natural phytoplankton community. This study revealed the environmental adaptation of the various sizes of phytoplankton community as well as their physiological response in the Arctic Beaufort Sea.

Can pelagic ciliates indicate vertical variation in the water quality status of western Arctic pelagic ecosystems?

The vertical pattern of pelagic ciliate communities was observed at eight layers in the Chukchi Sea and the northern Bering Sea of the western Arctic Ocean during the summer sea-ice reduction period (August 5 to August 24, 2016). A total of 44 ciliate species were identified, with seven species dominated the communities in the water column. Multivariate and univariate analyses demonstrated that: (1) community structures of ciliates vary significantly among eight water depths; (2) variations in the vertical distribution of ciliates were significantly correlated with changes in physicochemical variables, especially the ammonia; (3) the distributions of the three dominant species were significantly and positively related to the chlorophyll a and ammonia concentrations; and (4) species richness and abundance were significantly and positively correlated with the concentrations of ammonia and chlorophyll a. These results suggest that pelagic ciliates may reflect vertical variations in the water quality status of western Arctic ecosystems.