Constraints on crustal compositional architecture across the North China-Altaids transition and implications for craton margin reworking.

The phase of secular evolution of continents is manifested as the degree of compositional differentiation, modification and maturation of continental crusts, which is vital in understanding the mechanism of continental evolution but is difficult to quantify. Here we use integrated passive- and active-source seismic profiling to conduct joint analysis and inversion and derive Vs and Vp/Vs section models across the North China Craton (NCC) to southeastern Altaids boundary zone that bears a tectonic transition from a reworked ancient craton margin to a Phanerozoic accretionary orogen. We systematically exploited the imaged multiple physical properties as precise and delicate proxies to constrain the compositional architecture in the crust across this important tectonic transition subject to various crustal evolutional phases. Our Vs and Vp/Vs imaging, together with the existing isotopic data, characterizes the Yin Shan-Yan Shan belt as the northern NCC margin with layered homogeneous compositions that point to an evolved crust. However, the lower-crustal low-Vs/high-Vp/Vs signature that overlaps the shallowly dipping to horizontal reflective fabrics suggests that the crust of the northern NCC margin has undergone considerable reworking through lower-crustal-stretching-assisted melt migration and mixing since the late Paleozoic to Mesozoic eras. The process probably involved crust-mantle interaction and thus resulted in a compositionally modified ancient crustal basement. On the contrary, the southeastern Altaids domain manifests crustal complexity in compositions and structures inferred to be indicative of a juvenile crust of the Phanerozoic accretionary orogen. Our results provide deep physical-property constraints that shed new light on the crustal evolution of a complex craton margin.

Learning to Detect Salient Object With Multi-Source Weak Supervision.

High-cost pixel-level annotations makes it appealing to train saliency detection models with weak supervision. However, a single weak supervision source hardly contain enough information to train a well-performing model. To this end, we introduce a unified two-stage framework to learn from category labels, captions, web images and unlabeled images. In the first stage, we design a classification network (CNet) and a caption generation network (PNet), which learn to predict object categories and generate captions, respectively, meanwhile highlights the potential foreground regions. We present an attention transfer loss to transmit supervisions between two tasks and an attention coherence loss to encourage the networks to detect generally salient regions instead of task-specific regions. In the second stage, we create two complementary training datasets using CNet and PNet, i.e., natural image dataset with noisy labels for adapting saliency prediction network (SNet) to natural image input, and synthesized image dataset by pasting objects on background images for providing SNet with accurate ground-truth. During the testing phases, we only need SNet to predict saliency maps. Experiments indicate the performance of our method compares favorably against unsupervised, weakly supervised methods and even some supervised methods.

The occurrence and transmission characteristics of airborne H9N2 avian influenza virus.

To better understand the transmission route of H9N2 avian influenza virus (AIV), two duplicate trials were conducted to observe the process of aerosol infection and direct contact in specific pathogen free chickens. Fifteen chickens (G1) were inoculated with H9N2 AIV and housed together with another 15 chickens (G2) in the same positive-negative-pressure isolator (A). Fifteen chickens (G3) were bred in another isolator (B) which was connected with A so that air could flow unidirectionally from A to B. Air, oropharyngeal and cloacal swabs, and blood samples were collected for the detection of aerosolized virus, virus shedding, and seroconversion. AIV aerosols were initially detected at day 2-3 post inoculation (dpi), reaching peak concentrations at 7 dpi. Virus shedding was detected in all chickens of G2, but only in a part in G3 (T1: 87%, T2: 80%). Antibodies were initially detected at 4-5 dpi, peaking at 14-21 dpi. The results showed that H9N2 AIV could be transmitted by both aerosol exposure and direct contact.