The spatial characteristics of cracks are significant indicators to assess and evaluate the health of existing buildings and infrastructures. However, the current manual crack description method is time consuming and labor consuming. To improve the efficiency of crack inspection, advanced computer vision‐based techniques have been utilized to detect cracks automatically at image level and grid‐cell level. But existing crack detections are of (high specificity) low generality and inefficient, in terms that conventional approaches are unable to identify and measure diverse cracks concurrently at pixel level. Therefore, this research implements a novel deep learning technique named fully convolutional network (FCN) to address this problem. First, FCN is trained by feeding multiple types of cracks to semantically identify and segment pixel‐wise cracks at different scales. Then, the predicted crack segmentations are represented by single‐pixel width skeletons to quantitatively measure the morphological features of cracks, providing valuable crack indicators for assessment in practice, such as crack topology, crack length, max width, and mean width. To validate the prediction, the predicted segmentations are compared with recent advanced method for crack recognition and ground truth. For crack segmentation, the accuracy, precision, recall, and F1 score are 97.96%, 81.73%, 78.97%, and 79.95%, respectively. For crack length, the relative measurement error varies from −48.03% to 177.79%, meanwhile that ranges from −13.27% to 24.01% for crack width. The results show that FCN is feasible and sufficient for crack identification and measurement. Although the accuracy is not as high as CrackNet because of three types of errors, the prediction has been increased to pixel level and the training time has been dramatically decreased to several per cents of previous methods due to the novel end‐to‐end structure of FCN, which combines typical convolutional neural networks and deconvolutional layers.

No abstract

No abstract

COBE has provided us with a whole-sky map of the CBR anisotropies. However, even if the noise level is negligible when the four year COBE data are available, the cosmic variance will prevent us from obtaining information about the Gaussian nature of the primordial fluctuations. This important issue is addressed here by studying the angular bispectrum of the cosmic microwave background anisotropies. A general form of the angular bispectrum is given and the cosmic variance of the angular bispectrum for Gaussian fluctuations is calculated. The advantage of using the angular bispectrum is that one can choose to use the multipole moments which minimize the cosmic variance term. The non-Gaussian signals in most physically motivated non-Gaussian models are small compared with cosmic variance. Unless the amplitudes are large, the non-Gaussian signals are only detectable in the COBE data in those models where the angular bispectrum is flat or increases with increasing multipole moment.

Two G-quadruplex ligands [Pt(L(a))(DMSO)Cl] (Pt1) and [Pt(L(b))(DMSO)Cl] (Pt2) have been synthesized and fully characterized. The two complexes are more selective for SK-OV-3/DDP tumor cells versus normal cells (HL-7702). It was found that both Pt1 and Pt2 could be a telomerase inhibitor targeting G-quadruplex DNA. This is the first report demonstrating that telomeric, c-myc, and bcl-2 G-quadruplexes and caspase-3/9 preferred to bind with Pt2 rather than Pt1, which also can induce senescence and apoptosis. The different biological behavior of Pt1 and Pt2 may correlate with the presence of a 6-hydroxyl group in L(b). Importantly, Pt1 and Pt2 exhibited higher safety in vivo and more effective inhibitory effects on tumor growth in the HCT-8 and NCI-H460 xenograft mouse model, compared with cisplatin. Taken together, these mechanistic insights indicate that both Pt1 and Pt2 display low toxicity and could be novel anticancer drug candidates.