Attributes learning network for generalized zero-shot learning.

In the absence of unseen training data, zero-shot learning algorithms utilize the semantic knowledge shared by the seen and unseen classes to establish the connection between the visual space and the semantic space, so as to realize the recognition of the unseen classes. However, in real applications, the original semantic representation cannot well characterize both the class-specificity structure and discriminative information in dimension space, which leads to unseen classes being easily misclassified into seen classes. To tackle this problem, we propose a Salient Attributes Learning Network (SALN) to generate discriminative and expressive semantic representation under the supervision of the visual features. Meanwhile, ℓ1,2-norm constraint is employed to make the learned semantic representation well characterize the class-specificity structure and discriminative information in dimension space. Then feature alignment network projects the learned semantic representation into visual space and a relation network is adopted for classification. The performance of the proposed approach has made progress on the five benchmark datasets in generalized zero-shot learning task, and in-depth experiments indicate the effectiveness and excellence of our method.

Cross-modal distribution alignment embedding network for generalized zero-shot learning.

Many approaches in generalized zero-shot learning (GZSL) rely on cross-modal mapping between the image feature space and the class embedding space, which achieves knowledge transfer from seen to unseen classes. However, these two spaces are completely different space and their manifolds are inconsistent, the existing methods suffer from highly overlapped semantic description of different classes, as in GZSL tasks unseen classes can be easily misclassified into seen classes. To handle these problems, we adopt a novel semantic embedding network which helps to encode more discriminative information from initial semantic attributes to semantic embeddings in visual space. Meanwhile, a distribution alignment constraint is adopted to help keep the distribution of the learned semantic embeddings consistent with the distribution of real image features. Moreover, an auxiliary classifier is adopted to strengthen the quality of the learned semantic embeddings. Finally, a relation network is used to classify the unseen images by computing the relation scores between the semantic embeddings and image features, which is much more flexible than the fixed distance metric functions. Experimental results demonstrate that our proposed method is superior to other state-of-the-arts.

A TiS2/Celgard separator as an efficient polysulfide shuttling inhibitor for high-performance lithium-sulfur batteries.

The rapid capacity loss caused by the shuttling effect of polysulfides is one of the great challenges of Li-S batteries. In this work, we adopted a simple solid-phase sintering method to synthesize titanium disulfide (TiS2) and further demonstrated it as a superior modifier of separators for Li-S batteries. Two commonly adopted modification processes of separators, including vacuum filtration (VF) and slurry casting (SC) have been used to prepare TiS2/Celgard separators. TiS2-VF/Celgard can better restrain the polysulfide shuttling effect compared with TiS2-SC/Celgard. A TiS2-VF/Celgard-based Li-S battery has a reversible capacity of 771.6 mA h g-1, with a capacity retention of 645.6 mA h g-1 after 500 cycles at 2.0 C, corresponding to a capacity fading rate of ∼0.033% per cycle. This study has shown the potential of TiS2 as a multifunctional modifier of separators for high performance and long cycle life Li-S batteries.

One-Step Synthesis of Monodispersed Mesoporous Carbon Nanospheres for High-Performance Flexible Quasi-Solid-State Micro-Supercapacitors.

Cost-effective synthesis of carbon nanospheres with a desirable mesoporous network for diversified energy storage applications remains a challenge. Herein, a direct templating strategy is developed to fabricate monodispersed N-doped mesoporous carbon nanospheres (NMCSs) with an average particle size of 100 nm, a pore diameter of 4 nm, and a specific area of 1093 m2 g-1 . Hexadecyl trimethyl ammonium bromide and tetraethyl orthosilicate not only play key roles in the evolution of mesopores but also guide the assembly of phenolic resins to generate carbon nanospheres. Benefiting from the high surface area and optimum mesopore structure, NMCSs deliver a large specific capacitance up to 433 F g-1 in 1 m H2 SO4 . The NMCS electrodes-based symmetric sandwich supercapacitor has an output voltage of 1.4 V in polyvinyl alcohol/H2 SO4 gel electrolyte and delivers an energy density of 10.9 Wh kg-1 at a power density of 14014.5 W kg-1 . Notably, NMCSs can be directly applied through the mask-assisted casting technique by a doctor blade to fabricate micro-supercapacitors. The micro-supercapacitors exhibit excellent mechanical flexibility, long-term stability, and reliable power output.

Co2P@N,P-Codoped Carbon Nanofiber as a Free-Standing Air Electrode for Zn-Air Batteries: Synergy Effects of CoNx Satellite Shells.

Here, a free-standing electrode composed of cobalt phosphides (Co2P) supported by cobalt nitride moieties (CoNx) and an N,P-codoped porous carbon nanofiber (CNF) in one-step electrospinning of environmentally friendly benign phosphorous precursors is reported. Physiochemical characterization revealed the symbiotic relationship between a Co2P crystal and surrounding nanometer-sized CoNx moieties embedded in an N,P-codoped porous carbon matrix. Co2P@CNF shows high oxygen reduction reaction and oxygen evolution reaction performance owing to the synergistic effect of Co2P nanocrystals and the neighboring CoNx moieties, which have the optimum binding strength of reactants and facilitate the mass transfer. The free-standing Co2P@CNF air-cathode-based Zn-air batteries deliver a power density of 121 mW cm-2 at a voltage of 0.76 V. The overall overpotential of Co2P@CNF-based Zn-air batteries can be significantly reduced, with low discharge-charge voltage gap (0.81 V at 10 mA cm-2) and high cycling stability, which outperform the benchmark Pt/C-based Zn-air batteries. The one-step electrospinning method can serve as a universal platform to develop other high-performance transition-metal phosphide catalysts benefitting from the synergy effect of transition nitride satellite shells. The free-standing and flexible properties of Co2P@CNF make it a potential candidate for wearable electronic devices.