The key to fine-grained aircraft recognition is discovering the subtle traits that can distinguish different subcategories. Early approaches leverage part annotations of fine-grained objects to derive rich representations. However, manual labeling part information is cumbersome. In response to this issue, previous CNN-based methods reuse the backbone network to extract part-discrimination features, the inference process of which consumes much time. Therefore, we introduce generalized multiple instance learning (MIL) into fine-grained recognition. In generalized MIL, an aircraft is assumed to consist of multiple instances (such as head, tail, and body). Firstly, instance-level representations are obtained by the feature extractor and instance conversion component. Secondly, the obtained instance features are scored by an MIL classifier, which can yield high-level part semantics. Finally, a fine-grained object label is inferred by a MIL pooling function that aggregates multiple instance scores. The proposed approach is trained end-to-end without part annotations and complex location networks. Experimental evidence is conducted to prove the feasibility and effectiveness of our approach on combined aircraft images (CAIs).

A new eco-friendly process to comprehensively utilize boron-concentrate is proposed in this paper. In this process, Iron-concentrate, silicon dioxide, magn

As part of concepts for chemical energy storage of excess chemical energy produced from renewable sources, we investigated the performance of CuO/ZnO catalysts supported on mesoporous carbon to convert CO₂ hydrogenation to methanol. In this work, mesoporous carbon was used as the catalyst support for CuO-ZnO catalysts. Four catalysts with different precipitated precursors were synthesized and analyzed by N₂-physisorption, X-ray diffraction (XRD), thermogravimetric analysis (TG-DTG), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The results show that catalyst CZ-in situ had the highest turnover frequency (TOF) (2.8 × 10⁻³ s⁻¹) and methanol production rate (0.8 mmol g⁻¹·h⁻¹). The catalysts for co-precipitation of copper and zinc on carbon precursors are more active. Cu/ZnO domains that are accessible to the reactant gas are another reason for the catalysts being active. The Cu-ZnO interface is crucial to methanol catalyst activity.

Chemical looping partial oxidation of methane (CLPOM) is a low energy consumption and environmentally friendly new technology that can generate syngas. The main challenge is to find suitable oxygen carriers, which should be highly active, stable, low cost, and eco-friendly. This study found that Fe₂SiO₄ had good reactivity in the CLPOM process. Thermodynamic calculations were carried out by FactSage8.1 to demonstrate the feasibility of Fe₂SiO₄ as an oxygen carrier for CLPOM. Fe₂SiO₄ was prepared by the direct ball milling method and the high-temperature solid-phase synthesis method. The reaction properties of Fe₂SiO₄ were investigated in the fixed bed reactor. The XRD and FTIR results indicate that Fe₂SiO₄ can be synthesized successfully through the high-temperature solid-phase synthesis method. The results of fixed bed experiments showed that when the reaction temperature was 980 °C and the reaction time was 28 min, the $X_{C{H}_4}$ reached 87%, and the $S_{H_2}$ and $S_{\mathit{CO}}$ were 70% and 71%, respectively. Subsequently, 20 redox cycle experiments were conducted under the optimal reaction conditions. The results showed that Fe₂SiO₄ exhibited good reactivity in the first two cycles, and as the reaction progressed, the reduced oxygen carrier could not regain the lattice oxygen, leading to a decline in cyclic performance. This study demonstrates that Fe₂SiO₄ can couple CO₂ and CH₄ to produce syngas and is conducive to reducing carbon emissions.