Search

Underwater object detection is a prerequisite for underwater robots to achieve autonomous operation and ocean exploration. However, poor imaging quality, harsh underwater environments and concealed underwater targets greatly aggravate the difficulty of underwater object detection. In order to reduce underwater background interference and improve underwater object perception, we propose a multi-dimensional, multi-functional and multi-level attention module (mDFLAM). The multi-dimensional strategy first enhances the robustness of attention application by collecting valuable information in different target dimensions. The multi-functional strategy further improves the flexibility of attention calibration by capturing the importance of channel semantic information and the dependence of spatial location information. The multi-level strategy finally enriches the diversity of attention perception by extracting the intrinsic information under different receptive fields. In pre-processing and post-processing stages, cross-splitting and cross-linking stimulate the synergistic calibration advantage of multi-dimensional and multi-functional attention by redistributing channel dimensions and restoring feature states. In the attention calibration stage, adaptive fusion stimulates the synergistic calibration advantage of multi-level attention by assigning learnable parameters. In order to meet the high-precision and real-time requirements for underwater object detection, we integrate the plug-and-play mDFLAM into YOLO detectors. The full-port embedding further strengthens the semantic information expression by improving the feature fusion quality between scales. In underwater detection tasks, ablation and comparison experiments demonstrate the rationality and effectiveness of our attention design. In other detection tasks, our work shows good robustness and generalization. [ABSTRACT FROM AUTHOR]

The principal aim of present study was to assess the therapeutic efficacy of bone morphogenetic protein-7 (BMP-7) induced differentiation of bone marrow mesenchymal stem cells (BMSCs) in a rat acute spinal cord injury (SCI) model. BMSCs were isolated from rats, and then divided into a control and a BMP-7 induction groups. The proliferation ability of BMSCs and glial cell markers were determined. Forty Sprague–Dawley (SD) rats were randomly divided into sham, SCI, BMSC, and BMP7 + BMSC groups (n = 10). Among these rats, the recovery of hind limb motor function, the pathological related markers, and motor evoked potentials (MEP) were identified. BMSCs differentiated into neuron-like cells after the introduction of exogenous BMP-7. Interestingly, the expression levels of MAP-2 and Nestin increased, whereas the expression level of GFAP decreased after the treatment with exogenous BMP-7. Furthermore, the Basso, Beattie, and Bresnahan (BBB) score reached 19.33 ± 0.58 in the BMP-7 + BMSC group at day 42. Nissl bodies in the model group were reduced compared to the sham group. After 42 days, in both the BMSC and BMP-7 + BMSC groups, the number of Nissl bodies increased. This is especially so for the number of Nissl bodies in the BMP-7 + BMSC group, which was more than that in the BMSC group. The expression of Tuj-1 and MBP in BMP-7 + BMSC group increased, whereas the expression of GFAP decreased. Moreover, the MEP waveform decreased significantly after surgery. Furthermore, the waveform was wider and the amplitude was higher in BMP-7 + BMSC group than that in BMSC group. BMP-7 promotes BMSC proliferation, induces the differentiation of BMSCsinto neuron-like cells, and inhibits the formation of glial scar. BMP-7 plays a confident role in the recovery of SCI rats. [ABSTRACT FROM AUTHOR]

Marine object detection has become increasingly important in intelligent underwater robot. Because of color cast and blur in underwater images, features directly extracted from backbone networks usually lack interesting and discriminative characters, that affects performance on marine object detection. To this end, this paper proposes a novel refined marine object detector with attention-based spatial pyramid pooling networks and bidirectional feature fusion strategy to relieve the weakening of features and address marine object detection issues. Firstly, an attention-based spatial pyramid pooling network named as SA-SPPN is proposed to enrich interesting information and extend receptive field on original features extracted from backbone network. Based on enhanced multiple level features, the bidirectional feature fusion strategy is designed to fuse different level features and generate robust feature maps for detection. Specifically, the top-down connection could transfer semantic information from high-level features to enhance low-level features. The bottom-up pathway could extend resolution of high-level features. Furthermore, the cross-layer connections are integrated into both top-down passway and bottom-up passway to carry out multiple branch fusion. On bounding boxes regression phase, the distance-IoU loss is adopted to improve regression speed and accuracy. Finally, this paper conducts series experiments on underwater image datasets and URPC datasets to detect marine objects. The experimental results reveal that our approach could achieve impressive performance and reach 79.64% mAP on underwater image datasets, 79.31% mAP on URPC2019 datasets and 79.93% mAP on URPC2020 datasets, respectively. For standard object detection, the proposed algorithm also could realize notable performance and get 81.9% mAP on PASCAL VOC datasets. [ABSTRACT FROM AUTHOR]

The use of ceramic–metal laminated composites is an efficient way to improve the toughness of brittle ceramics. Uniform multilayered Al2O3–Ni laminated composites were prepared through a combined method of gel casting and hot pressing, which is less cumbersome than the traditional sequential lamination method. The wettability, diffusion and chemical reaction between Al2O3 and Ni were controlled through surface oxidation treatment of the Ni sheets. The results show that when the surface-oxidized Ni sheets were used, the microstructure of the Al2O3–Ni interface changed with the hot-pressing temperature. As the hot-pressing temperature was increased from 1150 to 1300 °C, a straight and clean interface formed, the interface became interlocked and, finally, a stable layer of NiAl2O4 formed at the interface. The interlocked microstructure was especially beneficial for interfacial bonding, and the plastic deformation of the Ni layers was restrained, thereby leading to slip steps rather than dimple formation on the fracture surfaces of the Ni layers. In addition, the toughening effects of plastic deformation and crack deflection in the Al2O3–Ni laminated composites were significant, which resulted in ductile fracture rather than brittle fracture. As a result, the Al2O3–Ni laminated composite that was prepared using the oxidized Ni sheets and the combined method of gel casting and hot pressing achieved a higher work of fracture than laminated composites that were prepared via other methods. [ABSTRACT FROM AUTHOR]

The east–west component of the magnetic field, Bϕ, as observed in solar magnetograms containing quiet regions, is found to change its sign when the regions cross the central meridian. It is seen in both HMI/SDO and VSM/SOLIS full disk vector magnetograms. A mismatch between the calibrated line-of-sight and transverse fields is the reason for this hemispheric bias problem. Here mismatch means that one of the fields is either over-estimated or under-estimated. For HMI data, the transverse field is over-estimated. This mismatch is caused ultimately by a filling factor that is not precisely determined when unresolved structures are present. An updated inversion procedure for HMI observations, developed recently, is able to derive the filling factor with reasonable accuracy. The new data show that the hemispheric bias problem has been mitigated substantially. [ABSTRACT FROM AUTHOR]

Turnip is a member of the Brassica rapa species and is characterized by a swollen taproot that develops from the hypocotyl and part of the root. Gibberellins (GAs) are plant growth regulators involved in promoting cell elongation and play important roles in many aspects of plant growth and development. Interestingly, exogenous application of GA3 was found to significantly inhibit taproot formation in turnip. Moreover, endogenous GA contents decreased during the early developmental stages of taproot formation, suggesting that GA plays a negative role in taproot formation. We examined the anatomical structure of the taproot and found that lignification of the xylem cell wall was enhanced after treatment with GA3. Yeast two-hybrid assays suggested the occurrence of protein interactions between DELLAs and NACs in turnip. We also found that the expression of NAC-targeted genes involved in lignification of the secondary cell wall was significantly upregulated upon GA3 treatment. Taken together, these results supported the hypothesis that GA induced DELLA proteins degradation to release NAC proteins and induced xylem lignification, therefore inhibiting taproot formation, providing new insight into the molecular mechanism underlying turnip taproot formation. [ABSTRACT FROM AUTHOR]

The safety of milk powder is closely related to everyone's life. In the production of milk powder, it is often easy to mix in foreign bodies, such as light and thin plastic, insects, metals, and other foreign bodies. In this paper, THz spectroscopy is used to detect common foreign bodies in milk powder, such as polymer materials (i.e., PP, PVC, and PE), insects, and metal gaskets. The detection of foreign bodies in milk powder can be realized by using THz spectroscopy combined with random forest discrimination method, and the recognition accuracy rate reaches 100%. The accuracy rate also reaches 100% when the types of foreign bodies are distinguished, thus, successfully realizing the identification and qualitative discrimination of foreign bodies in milk powder. At the same time, THz imaging technology is used to detect milk powder with foreign bodies. The outline and the boundaries of foreign bodies in milk powder are clear in the imaging from 0.5 to 1.0 THz. Thus, the size and shape of foreign bodies in milk powder can be well distinguished. This experiment verifies the feasibility of qualitative detection of foreign bodies in milk powder by the combination of terahertz imaging and spectrum, and established a better random forest detection model, which proposes a solution to ensure the safety of milk powder. [ABSTRACT FROM AUTHOR]