Search

Camouflaged object detection (COD) aims to identify the objects that conceal themselves in natural scenes. Accurate COD suffers from a number of challenges associated with low boundary contrast and the large variation of object appearances, e.g., object size and shape. To address these challenges, we propose a novel Context-aware Cross-level Fusion Network (C2F-Net), which fuses context-aware cross-level features for accurately identifying camouflaged objects. Specifically, we compute informative attention coefficients from multi-level features with our Attention-induced Cross-level Fusion Module (ACFM), which further integrates the features under the guidance of attention coefficients. We then propose a Dual-branch Global Context Module (DGCM) to refine the fused features for informative feature representations by exploiting rich global context information. Multiple ACFMs and DGCMs are integrated in a cascaded manner for generating a coarse prediction from high-level features. The coarse prediction acts as an attention map to refine the low-level features before passing them to our Camouflage Inference Module (CIM) to generate the final prediction. We perform extensive experiments on three widely used benchmark datasets and compare C2F-Net with state-of-the-art (SOTA) models. The results show that C2F-Net is an effective COD model and outperforms SOTA models remarkably. Further, an evaluation on polyp segmentation datasets demonstrates the promising potentials of our C2F-Net in COD downstream applications. Our code is publicly available at: https://github.com/Ben57882/C2FNet-TSCVT.

In this paper, we consider multiple solar-powered wireless nodes which utilize the harvested solar energy to transmit collected data to multiple unmanned aerial vehicles (UAVs) in the uplink. In this context, we jointly design UAV flight trajectories, UAV-node communication associations, and uplink power control to effectively utilize the harvested energy and manage co-channel interference within a finite time horizon. To ensure the fairness of wireless nodes, the design goal is to maximize the worst user rate. The joint design problem is highly non-convex and requires causal (future) knowledge of the instantaneous energy state information (ESI) and channel state information (CSI), which are difficult to predict in reality. To overcome these challenges, we propose an offline method based on convex optimization that only utilizes the average ESI and CSI. The problem is solved by three convex subproblems with successive convex approximation (SCA) and alternative optimization. We further design an online convex-assisted reinforcement learning (CARL) method to improve the system performance based on real-time environmental information. An idea of multi-UAV regulated flight corridors, based on the optimal offline UAV trajectories, is proposed to avoid unnecessary flight exploration by UAVs and enables us to improve the learning efficiency and system performance, as compared with the conventional reinforcement learning (RL) method. Computer simulations are used to verify the effectiveness of the proposed methods. The proposed CARL method provides 25% and 12% improvement on the worst user rate over the offline and conventional RL methods.

Although spatial indexes shorten the query response time, they rely on complex tree structures to narrow down the search space. Such structures in turn yield additional storage overhead and take a toll on index maintenance. Recently, there have been a flurry of efforts attempting to leverage Machine-Learning (ML) models to simplify the index structures. However, existing geospatial indexes can only index point data rather than complex geometries such as polygons and trajectories that are widely available in geospatial data. As a result, they cannot efficiently and correctly answer geometry relationship queries. This paper introduces GLIN, an indexing mechanism for spatial relationship queries on complex geometries. To achieve that, GLIN transforms geometries to Z-address intervals, and then harnesses an existing order-preserving learned index to model the cumulative distribution function between these intervals and the record positions. The lightweight learned index greatly reduces indexing overhead and provides faster or comparable query latency. Most importantly, GLIN augments spatial query windows to support queries exactly for common spatial relationships. Our experiments on real-world and synthetic datasets show that GLIN has 80\%-90\% lower storage overhead than Quad-Tree and 60% - 80% than R-tree and 30% - 70% faster query on medium selectivity. Moreover, GLIN's maintenance throughput is 1.5 times higher on insertion and 3 - 5 times higher on deletion.

Recent experimental discoveries of two-dimensional (2D) magnets have triggered intense research activities to search for atomically thin magnetic systems. Using first-principles calculations, we predict the emergence of 2D magnetism in the monolayers (MLs), few layers, and surfaces of nonmagnetic layered electrides Hf2X (X = S, Se, Te) consisting of three-atom-thick Hf-X-Hf stacks. It is revealed that each bulk Hf2X hosts a novel quantum state of Dirac nodal lines with a high density of states arising from Hf-5d cationic and interlayer anionic electrons around -0.9 eV below the Fermi level EF. However, for the MLs, few layers, and surfaces of Hf2X, such hybridized states are shifted toward EF to generate van Hove singularities, leading to a Stoner instability. The resulting surface ferromagnetism gives rise to strongly spin-polarized topological surface states at Hf2X(001), demonstrating that anionic electrons, 2D magnetism, and band topology are entangled with each other. Our findings will open new perspectives for the discovery of 2D magnets via exploiting surface effects in nonmagnetic layered electrides.

Charge and spin order is intimately related to superconductivity in copper oxide superconductors. To elucidate the competing orders in various nickel oxide compounds are crucial given the fact that superconductivity has been discovered in Nd$_{0.8}$Sr$_{0.2}$NiO$_2$ films. Herein, we report structural, electronic transport, magnetic, and thermodynamic characterizations on single crystals of La$_3$Ni$_2$O$_7$ and La$_3$Ni$_2$O$_6$. La$_3$Ni$_2$O$_7$ is metallic with mixed Ni$^{2+}$ and Ni$^{3+}$ valent states. Resistivity measurements yield two transition-like kinks at $\sim$110 and 153 K, respectively. The kink at 153 K is further revealed from magnetization and specific heat measurements, indicative of the formation of charge and spin order. La$_3$Ni$_2$O$_6$ single crystals obtained from topochemical reduction of La$_3$Ni$_2$O$_7$ are insulating and show an anomaly at $\sim$176 K on magnetic susceptibility. The transition-like behaviors of La$_3$Ni$_2$O$_7$ and La$_3$Ni$_2$O$_6$ are analogous to the charge and spin order observed in La$_4$Ni$_3$O$_{10}$ and La$_4$Ni$_3$O$_8$, suggesting charge and spin order is a common feature in the ternary La-Ni-O system with mixed-valent states of nickel., Comment: 7 pages, 4 figures, 2 tables

We report the path from the charge density wave (CDW)-bearing superconductor CuIr2Te4 to the metal insulator transition (MIT)-bearing compound CuIr2S4 by chemical alloying with the gradual substitution of S for Te. The evolution of structural and physical properties of the CuIr2Te4-xSx polycrystalline system with the doping range from 0 to 4 is systemically examined. The X-ray diffraction (XRD) results imply CuIr2Te4-xSx in the range between 0 and 0.5 crystallizes in a NiAs defected trigonal structure, whereas it adapts to the cubic spinel structure in the doping range from 3.6 to 4 and it is a mixed phase in the doping range from 0.5 to 3.6. Unexpectedly, the resistivity and magnetization measurements reveal that small-concentration S substitution for Te can suppress the CDW transition, but it reappears around x = 0.2, and the CDW transition temperature enhances clearly as x augments in the range from 0.2 to 0.5. Besides, the superconducting critical temperature (Tc) first increases with S doping content and then decreases after reaching a maximum Tc = 2.82 K for CuIr2Te3.85S0.15. MIT order has been observed in the spinel region from 3.6 to 4 associated with TMI increasing with x increasing. Finally, the rich electronic phase diagram of temperature versus x for this CuIr2Te4-xSx system is assembled, where the superconducting dome is associated with the suppression and re-emergence of CDW as well as MIT states at the end upon sulfur substitution in the CuIr2Te4-xSx chalcogenides., Comment: 33 pages, 8 figures, 4 tables; The manuscript with the same title will be published by JOURNAL OF PHYSICS-CONDENSED MATTER

Transformer-based language models are able to generate fluent text and be efficiently adapted across various natural language generation tasks. However, language models that are pretrained on large unlabeled web text corpora have been shown to suffer from degenerating toxic content and social bias behaviors, consequently hindering their safe deployment. Various detoxification methods were proposed to mitigate the language model's toxicity; however, these methods struggled to detoxify language models when conditioned on prompts that contain specific social identities related to gender, race, or religion. In this study, we propose Reinforce-Detoxify; A reinforcement learning-based method for mitigating toxicity in language models. We address the challenge of safety in language models and propose a new reward model that is able to detect toxic content and mitigate unintended bias towards social identities in toxicity prediction. The experiments demonstrate that the Reinforce-Detoxify method for language model detoxification outperforms existing detoxification approaches in automatic evaluation metrics, indicating the ability of our approach in language model detoxification and less prone to unintended bias toward social identities in generated content.