Showing papers by "Hong Kong University of Science and Technology published in 2019"

Federated Machine Learning: Concept and Applications

Abstract: Today’s artificial intelligence still faces two major challenges. One is that, in most industries, data exists in the form of isolated islands. The other is the strengthening of data privacy and security. We propose a possible solution to these challenges: secure federated learning. Beyond the federated-learning framework first proposed by Google in 2016, we introduce a comprehensive secure federated-learning framework, which includes horizontal federated learning, vertical federated learning, and federated transfer learning. We provide definitions, architectures, and applications for the federated-learning framework, and provide a comprehensive survey of existing works on this subject. In addition, we propose building data networks among organizations based on federated mechanisms as an effective solution to allowing knowledge to be shared without compromising user privacy.

Federated Machine Learning: Concept and Applications

Abstract: Today's AI still faces two major challenges. One is that in most industries, data exists in the form of isolated islands. The other is the strengthening of data privacy and security. We propose a possible solution to these challenges: secure federated learning. Beyond the federated learning framework first proposed by Google in 2016, we introduce a comprehensive secure federated learning framework, which includes horizontal federated learning, vertical federated learning and federated transfer learning. We provide definitions, architectures and applications for the federated learning framework, and provide a comprehensive survey of existing works on this subject. In addition, we propose building data networks among organizations based on federated mechanisms as an effective solution to allow knowledge to be shared without compromising user privacy.

The Roadmap to 6G: AI Empowered Wireless Networks

Abstract: The recent upsurge of diversified mobile applications, especially those supported by AI, is spurring heated discussions on the future evolution of wireless communications. While 5G is being deployed around the world, efforts from industry and academia have started to look beyond 5G and conceptualize 6G. We envision 6G to undergo an unprecedented transformation that will make it substantially different from the previous generations of wireless cellular systems. In particular, 6G will go beyond mobile Internet and will be required to support ubiquitous AI services from the core to the end devices of the network. Meanwhile, AI will play a critical role in designing and optimizing 6G architectures, protocols, and operations. In this article, we discuss potential technologies for 6G to enable mobile AI applications, as well as AI-enabled methodologies for 6G network design and optimization. Key trends in the evolution to 6G will also be discussed.

GCNet: Non-Local Networks Meet Squeeze-Excitation Networks and Beyond

Abstract: The Non-Local Network (NLNet) presents a pioneering approach for capturing long-range dependencies, via aggregating query-specific global context to each query position. However, through a rigorous empirical analysis, we have found that the global contexts modeled by non-local network are almost the same for different query positions within an image. In this paper, we take advantage of this finding to create a simplified network based on a query-independent formulation, which maintains the accuracy of NLNet but with significantly less computation. We further observe that this simplified design shares similar structure with Squeeze-Excitation Network (SENet). Hence we unify them into a three-step general framework for global context modeling. Within the general framework, we design a better instantiation, called the global context (GC) block, which is lightweight and can effectively model the global context. The lightweight property allows us to apply it for multiple layers in a backbone network to construct a global context network (GCNet), which generally outperforms both simplified NLNet and SENet on major benchmarks for various recognition tasks.

Generalizing from a Few Examples: A Survey on Few-Shot Learning

Abstract: Machine learning has been highly successful in data-intensive applications but is often hampered when the data set is small. Recently, Few-Shot Learning (FSL) is proposed to tackle this problem. Using prior knowledge, FSL can rapidly generalize to new tasks containing only a few samples with supervised information. In this paper, we conduct a thorough survey to fully understand FSL. Starting from a formal definition of FSL, we distinguish FSL from several relevant machine learning problems. We then point out that the core issue in FSL is that the empirical risk minimized is unreliable. Based on how prior knowledge can be used to handle this core issue, we categorize FSL methods from three perspectives: (i) data, which uses prior knowledge to augment the supervised experience; (ii) model, which uses prior knowledge to reduce the size of the hypothesis space; and (iii) algorithm, which uses prior knowledge to alter the search for the best hypothesis in the given hypothesis space. With this taxonomy, we review and discuss the pros and cons of each category. Promising directions, in the aspects of the FSL problem setups, techniques, applications and theories, are also proposed to provide insights for future research.

Cation and anion immobilization through chemical bonding enhancement with fluorides for stable halide perovskite solar cells

Abstract: Defects play an important role in the degradation processes of hybrid halide perovskite absorbers, impeding their application for solar cells. Among all defects, halide anion and organic cation vacancies are ubiquitous, promoting ion diffusion and leading to thin-film decomposition at surfaces and grain boundaries. Here, we employ fluoride to simultaneously passivate both anion and cation vacancies, by taking advantage of the extremely high electronegativity of fluoride. We obtain a power conversion efficiency of 21.46% (and a certified 21.3%-efficient cell) in a device based on the caesium, methylammonium (MA) and formamidinium (FA) triple-cation perovskite (Cs0.05FA0.54MA0.41)Pb(I0.98Br0.02)3 treated with sodium fluoride. The device retains 90% of its original power conversion efficiency after 1,000 h of operation at the maximum power point. With the help of first-principles density functional theory calculations, we argue that the fluoride ions suppress the formation of halide anion and organic cation vacancies, through a unique strengthening of the chemical bonds with the surrounding lead and organic cations. Defects and defect migration are detrimental for perovskite solar cell efficiency and long-term stability. Li et al. show that fluoride is able to suppress the formation of halide anion and organic cation vacancy defects by restraining the relative ions via ionic and hydrogen bonds.

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Abstract: In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications. Traditional cloudbased Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL

Spatiotemporal Multi-Graph Convolution Network for Ride-Hailing Demand Forecasting

Abstract: Region-level demand forecasting is an essential task in ridehailing services. Accurate ride-hailing demand forecasting can guide vehicle dispatching, improve vehicle utilization, reduce the wait-time, and mitigate traffic congestion. This task is challenging due to the complicated spatiotemporal dependencies among regions. Existing approaches mainly focus on modeling the Euclidean correlations among spatially adjacent regions while we observe that non-Euclidean pair-wise correlations among possibly distant regions are also critical for accurate forecasting. In this paper, we propose the spatiotemporal multi-graph convolution network (ST-MGCN), a novel deep learning model for ride-hailing demand forecasting. We first encode the non-Euclidean pair-wise correlations among regions into multiple graphs and then explicitly model these correlations using multi-graph convolution. To utilize the global contextual information in modeling the temporal correlation, we further propose contextual gated recurrent neural network which augments recurrent neural network with a contextual-aware gating mechanism to re-weights different historical observations. We evaluate the proposed model on two real-world large scale ride-hailing demand datasets and observe consistent improvement of more than 10% over stateof-the-art baselines.

Topological phases in acoustic and mechanical systems

Abstract: The study of classical wave physics has been reinvigorated by incorporating the concept of the geometric phase, which has its roots in optics, and topological notions that were previously explored in condensed matter physics. Recently, sound waves and a variety of mechanical systems have emerged as excellent platforms that exemplify the universality and diversity of topological phases. In this Review, we introduce the essential physical concepts that underpin various classes of topological phenomena realized in acoustic and mechanical systems: Dirac points, the quantum Hall, quantum spin Hall and valley Hall effects, Floquet topological phases, 3D gapless states and Weyl crystals. This Review describes topological phenomena that can be realized in acoustic and mechanical systems. Methods of symmetry breaking are described, along with the consequences and rich phenomena that emerge.

Stereo R-CNN Based 3D Object Detection for Autonomous Driving

Abstract: We propose a 3D object detection method for autonomous driving by fully exploiting the sparse and dense, semantic and geometry information in stereo imagery. Our method, called Stereo R-CNN, extends Faster R-CNN for stereo inputs to simultaneously detect and associate object in left and right images. We add extra branches after stereo Region Proposal Network (RPN) to predict sparse keypoints, viewpoints, and object dimensions, which are combined with 2D left-right boxes to calculate a coarse 3D object bounding box. We then recover the accurate 3D bounding box by a region-based photometric alignment using left and right RoIs. Our method does not require depth input and 3D position supervision, however, outperforms all existing fully supervised image-based methods. Experiments on the challenging KITTI dataset show that our method outperforms the state-of-the-art stereo-based method by around 30% AP on both 3D detection and 3D localization tasks. Code will be made publicly available.

Aggregation-induced emission: fundamental understanding and future developments

Abstract: Since the introduction of the concept of aggregation-induced emission (AIE) in 2001, many research groups have become involved in AIE research. Aggregation-induced emission luminogens (AIEgens) have emerged as a novel type of advanced material with excellent performance in various fields. Much effort has been devoted to determining the AIE mechanism(s) by theoreticians and experimentalists. Restriction of intramolecular motion has been recognized as the general working mechanism of AIE, but the mechanims of some AIE systems still remain unclear. In this focus article, the progress of the fundamental understanding of the AIE mechanism is reviewed and the future developments in AIE research are discussed. The goal is to provide a brief yet insightful introduction and interpretation of the subject to both new and experienced AIE researchers.

Strain engineering in perovskite solar cells and its impacts on carrier dynamics.

Abstract: The mixed halide perovskites have emerged as outstanding light absorbers for efficient solar cells. Unfortunately, it reveals inhomogeneity in these polycrystalline films due to composition separation, which leads to local lattice mismatches and emergent residual strains consequently. Thus far, the understanding of these residual strains and their effects on photovoltaic device performance is absent. Herein we study the evolution of residual strain over the films by depth-dependent grazing incident X-ray diffraction measurements. We identify the gradient distribution of in-plane strain component perpendicular to the substrate. Moreover, we reveal its impacts on the carrier dynamics over corresponding solar cells, which is stemmed from the strain induced energy bands bending of the perovskite absorber as indicated by first-principles calculations. Eventually, we modulate the status of residual strains in a controllable manner, which leads to enhanced PCEs up to 20.7% (certified) in devices via rational strain engineering. The residual strains in the mixed halide perovskite thin films and their effects on the solar cell devices are less understood. Here Zhu et al. study the impact of the gradient in-plane strain on the carrier dynamics of the strained perovskite films and optimize the device efficiency.

Insights into the Fate and Removal of Antibiotics in Engineered Biological Treatment Systems: A Critical Review

Abstract: Antibiotics, the most frequently prescribed drugs of modern medicine, are extensively used for both human and veterinary applications. Antibiotics from different wastewater sources (e.g., municipal, hospitals, animal production, and pharmaceutical industries) ultimately are discharged into wastewater treatment plants. Sorption and biodegradation are the two major removal pathways of antibiotics during biological wastewater treatment processes. This review provides the fundamental insights into sorption mechanisms and biodegradation pathways of different classes of antibiotics with diverse physical-chemical attributes. Important factors affecting sorption and biodegradation behavior of antibiotics are also highlighted. Furthermore, this review also sheds light on the critical role of extracellular polymeric substances on antibiotics adsorption and their removal in engineered biological wastewater treatment systems. Despite major advancements, engineered biological wastewater treatment systems are only moderately effective (48-77%) in the removal of antibiotics. In this review, we systematically summarize the behavior and removal of different antibiotics in various biological treatment systems with discussion on their removal efficiency, removal mechanisms, critical bioreactor operating conditions affecting antibiotics removal, and recent innovative advancements. Besides, relevant background information including antibiotics classification, physical-chemical properties, and their occurrence in the environment from different sources is also briefly covered. This review aims to advance our understanding of the fate of various classes of antibiotics in engineered biological wastewater treatment systems and outlines future research directions.

Earthquake-Induced Chains of Geologic Hazards: Patterns, Mechanisms, and Impacts

Abstract: Large earthquakes initiate chains of surface processes that last much longer than the brief moments of strong shaking. Most moderate‐ and large‐magnitude earthquakes trigger landslides, ranging from small failures in the soil cover to massive, devastating rock avalanches. Some landslides dam rivers and impound lakes, which can collapse days to centuries later, and flood mountain valleys for hundreds of kilometers downstream. Landslide deposits on slopes can remobilize during heavy rainfall and evolve into debris flows. Cracks and fractures can form and widen on mountain crests and flanks, promoting increased frequency of landslides that lasts for decades. More gradual impacts involve the flushing of excess debris downstream by rivers, which can generate bank erosion and floodplain accretion as well as channel avulsions that affect flooding frequency, settlements, ecosystems, and infrastructure. Ultimately, earthquake sequences and their geomorphic consequences alter mountain landscapes over both human and geologic time scales. Two recent events have attracted intense research into earthquake‐induced landslides and their consequences: the magnitude M 7.6 Chi‐Chi, Taiwan earthquake of 1999, and the M 7.9 Wenchuan, China earthquake of 2008. Using data and insights from these and several other earthquakes, we analyze how such events initiate processes that change mountain landscapes, highlight research gaps, and suggest pathways toward a more complete understanding of the seismic effects on the Earth's surface.

Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data

Abstract: Deep learning techniques for point cloud data have demonstrated great potentials in solving classical problems in 3D computer vision such as 3D object classification and segmentation. Several recent 3D object classification methods have reported state-of-the-art performance on CAD model datasets such as ModelNet40 with high accuracy (~92\%). Despite such impressive results, in this paper, we argue that object classification is still a challenging task when objects are framed with real-world settings. To prove this, we introduce ScanObjectNN, a new real-world point cloud object dataset based on scanned indoor scene data. From our comprehensive benchmark, we show that our dataset poses great challenges to existing point cloud classification techniques as objects from real-world scans are often cluttered with background and/or are partial due to occlusions. We identify three key open problems for point cloud object classification, and propose new point cloud classification neural networks that achieve state-of-the-art performance on classifying objects with cluttered background. Our dataset and code are publicly available in our project page https://hkust-vgd.github.io/scanobjectnn/.

Molecular Motion in Aggregates: Manipulating TICT for Boosting Photothermal Theranostics

Abstract: Planar donor and acceptor (D–A) conjugated structures are generally believed to be the standard for architecting highly efficient photothermal theranostic agents, in order to restrict intramolecular motions in aggregates (nanoparticles). However, other channels of extra nonradiative decay may be blocked. Now this challenge is addressed by proposing an “abnormal” strategy based on molecular motion in aggregates. Molecular rotors and bulky alkyl chains are grafted to the central D–A core to lower intermolecular interaction. The enhanced molecular motion favors the formation of a dark twisted intramolecular charge transfer state, whose nonradiative decay enhances the photothermal properties. Result shows that small-molecule NIRb14 with long alkyl chains branched at the second carbon exhibits enhanced photothermal properties compared with NIRb6, with short branched chains, and much higher than NIR6, with short linear chains, and the commercial gold nanorods. Both in vitro and in vivo experiments demonstrate t...

Plug and Play Language Models: A Simple Approach to Controlled Text Generation

Abstract: Large transformer-based language models (LMs) trained on huge text corpora have shown unparalleled generation capabilities. However, controlling attributes of the generated language (e.g. switching topic or sentiment) is difficult without modifying the model architecture or fine-tuning on attribute-specific data and entailing the significant cost of retraining. We propose a simple alternative: the Plug and Play Language Model (PPLM) for controllable language generation, which combines a pretrained LM with one or more simple attribute classifiers that guide text generation without any further training of the LM. In the canonical scenario we present, the attribute models are simple classifiers consisting of a user-specified bag of words or a single learned layer with 100,000 times fewer parameters than the LM. Sampling entails a forward and backward pass in which gradients from the attribute model push the LM's hidden activations and thus guide the generation. Model samples demonstrate control over a range of topics and sentiment styles, and extensive automated and human annotated evaluations show attribute alignment and fluency. PPLMs are flexible in that any combination of differentiable attribute models may be used to steer text generation, which will allow for diverse and creative applications beyond the examples given in this paper.

Response of microbial communities to biochar-amended soils: a critical review

Abstract: Application of biochar to soils changes soil physicochemical properties and stimulates the activities of soil microorganisms that influence soil quality and plant performance. Studying the response of soil microbial communities to biochar amendments is important for better understanding interactions of biochar with soil, as well as plants. However, the effect of biochar on soil microorganisms has received less attention than its influences on soil physicochemical properties. In this review, the following key questions are discussed: (i) how does biochar affect soil microbial activities, in particular soil carbon (C) mineralization, nutrient cycling, and enzyme activities? (ii) how do microorganisms respond to biochar amendment in contaminated soils? and (iii) what is the role of biochar as a growth promoter for soil microorganisms? Many studies have demonstrated that biochar-soil application enhances the soil microbial biomass with substantial changes in microbial community composition. Biochar amendment changes microbial habitats, directly or indirectly affects microbial metabolic activities, and modifies the soil microbial community in terms of their diversity and abundance. However, chemical properties of biochar, (especially pH and nutrient content), and physical properties such as pore size, pore volume, and specific surface area play significant roles in determining the efficacy of biochar on microbial performance as biochar provides suitable habitats for microorganisms. The mode of action of biochar leading to stimulation of microbial activities is complex and is influenced by the nature of biochar as well as soil conditions.

Protecting World Leaders Against Deep Fakes

Abstract: The creation of sophisticated fake videos has been largely relegated to Hollywood studios or state actors. Recent advances in deep learning, however, have made it significantly easier to create sophisticated and compelling fake videos. With relatively modest amounts of data and computing power, the average person can, for example, create a video of a world leader confessing to illegal activity leading to a constitutional crisis, a military leader saying something racially insensitive leading to civil unrest in an area of military activity, or a corporate titan claiming that their profits are weak leading to global stock manipulation. These so called deep fakes pose a significant threat to our democracy, national security, and society. To contend with this growing threat, we describe a forensic technique that models facial expressions and movements that typify an individual’s speaking pattern. Although not visually apparent, these correlations are often violated by the nature of how deep-fake videos are created and can, therefore, be used for authentication.

Gated Context Aggregation Network for Image Dehazing and Deraining

Abstract: Image dehazing aims to recover the uncorrupted content from a hazy image. Instead of leveraging traditional low-level or handcrafted image priors as the restoration constraints, e.g., dark channels and increased contrast, we propose an end-to-end gated context aggregation network to directly restore the final haze-free image. In this network, we adopt the latest smoothed dilation technique to help remove the gridding artifacts caused by the widely-used dilated convolution with negligible extra parameters, and leverage a gated sub-network to fuse the features from different levels. Extensive experiments demonstrate that our method can surpass previous state-of-the-art methods by a large margin both quantitatively and qualitatively. In addition, to demonstrate the generality of the proposed method, we further apply it to the image deraining task, which also achieves the state-of-the-art performance.

ShellNet: Efficient Point Cloud Convolutional Neural Networks Using Concentric Shells Statistics

Abstract: Deep learning with 3D data has progressed significantly since the introduction of convolutional neural networks that can handle point order ambiguity in point cloud data. While being able to achieve good accuracies in various scene understanding tasks, previous methods often have low training speed and complex network architecture. In this paper, we address these problems by proposing an efficient end-to-end permutation invariant convolution for point cloud deep learning. Our simple yet effective convolution operator named ShellConv uses statistics from concentric spherical shells to define representative features and resolve the point order ambiguity, allowing traditional convolution to perform on such features. Based on ShellConv we further build an efficient neural network named ShellNet to directly consume the point clouds with larger receptive fields while maintaining less layers. We demonstrate the efficacy of ShellNet by producing state-of-the-art results on object classification, object part segmentation, and semantic scene segmentation while keeping the network very fast to train.

A monothiophene unit incorporating both fluoro and ester substitution enabling high-performance donor polymers for non-fullerene solar cells with 16.4% efficiency

Abstract: Thiophene and its derivatives have been extensively used in organic electronics, particularly in the field of polymer solar cells (PSCs). Significant research efforts have been dedicated to modifying thiophene-based units by attaching electron-donating or withdrawing groups to tune the energy levels of conjugated materials. Herein, we report the design and synthesis of a novel thiophene derivative, FE-T, featuring a monothiophene functionalized with both an electron-withdrawing fluorine atom (F) and an ester group (E). The FE-T unit possesses distinctive advantages of both F and E groups, the synergistic effects of which enable significant downshifting of the energy levels and enhanced aggregation/crystallinity of the resulting organic materials. Shown in this work are a series of polymers obtained by incorporating the FE-T unit into a PM6 polymer to fine-tune the energetics and morphology of this high-performance PSC material. The optimal polymer in the series shows a downshifted HOMO and an improved morphology, leading to a high PCE of 16.4% with a small energy loss (0.53 eV) enabled by the reduced non-radiative energy loss (0.23 eV), which are among the best values reported for non-fullerene PSCs to date. This work shows that the FE-T unit is a promising building block to construct donor polymers for high-performance organic photovoltaic cells.

Ridesourcing systems: A framework and review

Abstract: With the rapid development and popularization of mobile and wireless communication technologies, ridesourcing companies have been able to leverage internet-based platforms to operate e-hailing services in many cities around the world. These companies connect passengers and drivers in real time and are disruptively changing the transportation industry. As pioneers in a general sharing economy context, ridesourcing shared transportation platforms consist of a typical two-sided market. On the demand side, passengers are sensitive to the price and quality of the service. On the supply side, drivers, as freelancers, make working decisions flexibly based on their income from the platform and many other factors. Diverse variables and factors in the system are strongly endogenous and interactively dependent. How to design and operate ridesourcing systems is vital—and challenging—for all stakeholders: passengers/users, drivers/service providers, platforms, policy makers, and the general public. In this paper, we propose a general framework to describe ridesourcing systems. This framework can aid understanding of the interactions between endogenous and exogenous variables, their changes in response to platforms’ operational strategies and decisions, multiple system objectives, and market equilibria in a dynamic manner. Under the proposed general framework, we summarize important research problems and the corresponding methodologies that have been and are being developed and implemented to address these problems. We conduct a comprehensive review of the literature on these problems in different areas from diverse perspectives, including (1) demand and pricing, (2) supply and incentives, (3) platform operations, and (4) competition, impacts, and regulations. The proposed framework and the review also suggest many avenues requiring future research.

CMFL: Mitigating Communication Overhead for Federated Learning

Abstract: Federated Learning enables mobile users to collaboratively learn a global prediction model by aggregating their individual updates without sharing the privacy-sensitive data. As mobile devices usually have limited data plan and slow network connections to the central server where the global model is maintained, mitigating the communication overhead is of paramount importance. While existing works mainly focus on reducing the total bits transferred in each update via data compression, we study an orthogonal approach that identifies irrelevant updates made by clients and precludes them from being uploaded for reduced network footprint. Following this idea, we propose Communication-Mitigated Federated Learning (CMFL) in this paper. CMFL provides clients with feedback information regarding the global tendency of model updating. Each client checks if its update aligns with this global tendency and is relevant enough to model improvement. By avoiding uploading those irrelevant updates to the server, CMFL can substantially reduce the communication overhead while still guaranteeing the learning convergence. CMFL is shown to achieve general improvement in communication efficiency for almost all of the existing federated learning schemes. We evaluate CMFL through extensive simulations and EC2 emulations. Compared with vanilla Federated Learning, CMFL yields 13.97x communication efficiency in terms of the reduction of network footprint. When applied to Federated Multi-Task Learning, CMFL improves the communication efficiency by 5.7x with 4% higher prediction accuracy.

Building ultraconformal protective layers on both secondary and primary particles of layered lithium transition metal oxide cathodes

Abstract: Despite their relatively high capacity, layered lithium transition metal oxides suffer from crystal and interfacial structural instability under aggressive electrochemical and thermal driving forces, leading to rapid performance degradation and severe safety concerns. Here we report a transformative approach using an oxidative chemical vapour deposition technique to build a protective conductive polymer (poly(3,4-ethylenedioxythiophene)) skin on layered oxide cathode materials. The ultraconformal poly(3,4-ethylenedioxythiophene) skin facilitates the transport of lithium ions and electrons, significantly suppresses the undesired layered to spinel/rock-salt phase transformation and the associated oxygen loss, mitigates intergranular and intragranular mechanical cracking, and effectively stabilizes the cathode–electrolyte interface. This approach remarkably enhances the capacity and thermal stability under high-voltage operation. Building a protective skin at both secondary and primary particle levels of layered oxides offers a promising design strategy for Ni-rich cathodes towards high-energy, long-life and safe lithium-ion batteries. Intensive research efforts are underway to enable applications of layered lithium transition metal oxides in batteries. Here the authors report an oxidative chemical vapour deposition technique to conformally coat both the primary and the secondary particles of these oxides to unleash potential applications.

The Roadmap to 6G -- AI Empowered Wireless Networks

Abstract: The recent upsurge of diversified mobile applications, especially those supported by Artificial Intelligence (AI), is spurring heated discussions on the future evolution of wireless communications. While 5G is being deployed around the world, efforts from industry and academia have started to look beyond 5G and conceptualize 6G. We envision 6G to undergo an unprecedented transformation that will make it substantially different from the previous generations of wireless cellular systems. In particular, 6G will go beyond mobile Internet and will be required to support ubiquitous AI services from the core to the end devices of the network. Meanwhile, AI will play a critical role in designing and optimizing 6G architectures, protocols, and operations. In this article, we discuss potential technologies for 6G to enable mobile AI applications, as well as AI-enabled methodologies for 6G network design and optimization. Key trends in the evolution to 6G will also be discussed.

Tightly Coupled 3D Lidar Inertial Odometry and Mapping

Abstract: Ego-motion estimation is a fundamental requirement for most mobile robotic applications. By sensor fusion, we can compensate the deficiencies of stand-alone sensors and provide more reliable estimations. We introduce a tightly coupled lidar-IMU fusion method in this paper. By jointly minimizing the cost derived from lidar and IMU measurements, the lidarIMU odometry (LIO) can perform well with considerable drifts after long-term experiment, even in challenging cases where the lidar measurement can be degraded. Besides, to obtain more reliable estimations of the lidar poses, a rotation-constrained refinement algorithm (LIO-mapping) is proposed to further align the lidar poses with the global map. The experiment results demonstrate that the proposed method can estimate the poses of the sensor pair at the IMU update rate with high precision, even under fast motion conditions or with insufficient features.

MetaPruning: Meta Learning for Automatic Neural Network Channel Pruning

Abstract: In this paper, we propose a novel meta learning approach for automatic channel pruning of very deep neural networks. We first train a PruningNet, a kind of meta network, which is able to generate weight parameters for any pruned structure given the target network. We use a simple stochastic structure sampling method for training the PruningNet. Then, we apply an evolutionary procedure to search for good-performing pruned networks. The search is highly efficient because the weights are directly generated by the trained PruningNet and we do not need any finetuning at search time. With a single PruningNet trained for the target network, we can search for various Pruned Networks under different constraints with little human participation. Compared to the state-of-the-art pruning methods, we have demonstrated superior performances on MobileNet V1/V2 and ResNet. Codes are available on https://github.com/liuzechun/MetaPruning.

Aggregation-Induced Emission Luminogens for Activity-Based Sensing.

Abstract: Fluorescent sensing has emerged as a powerful tool for detecting various analytes and visualizing numerous biological processes by virtue of its superb sensitivity, rapidness, excellent temporal resolution, easy operation, and low cost. Of particular interest is activity-based sensing (ABS), a burgeoning sensing approach that is actualized on the basis of dynamic molecular reactivity rather than conventional lock-and-key molecular recognition. ABS has been recognized to possess some distinct advantages, such as high specificity, extraordinary sensitivity, and accurate signal outputs. A majority of ABS sensors are constructed by modifying conventional fluorogens, which are strongly emissive when molecularly dissolved in solvents but experience emission quenching upon aggregate formation or concentration increase. The aggregation-caused quenching (ACQ) phenomenon leads to a limited amount of labeling of the analyte with the sensor and low photobleaching resistance, which could impede practical applications of the ABS protocol. As an anti-ACQ phenomenon, aggregation-induced emission (AIE) provides a straightforward solution to the ACQ problem. Thanks to their intrinsic advantages, including high photobleaching threshold, high signal-to-noise ratio, fluorescence turn-on nature, and large Stokes shift, AIE-active luminogens (AIEgens) represent a class of extraordinary fluorogen alternatives for the ABS protocol. The use of AIEgen-involved ABS can integrate the advantages of AIEgens and ABS, and additionally, the AIE process offers some unique properties to the ABS approach. For instance, in some cases of water-soluble AIEgen-involved ABS, chemical reaction not only leads to a chang in the emission color of the AIEgens but also causes solubility variations, which could result in specific "light-up" signaling. In this Account, the basic concepts and mechanistic insights of the ABS approach involving the AIE principle are briefly summarized, and then we highlight the new breakthroughs, seminal studies, and trends in the area that have been most recently reported by our group. This emerging sensing protocol has been successfully utilized for detecting an array of targets including ions, small molecules, biomacromolecules, and microenvironments, all of which closely relate to human health, medical, and public concerns. These detections are smoothly achieved on the basis of various reactions (e.g., hydrolysis, boronate cleavage, dephosphorylation, addition, cyclization, and rearrangement reactions) through different sensing principles. In these studies, the AIEgen-involved ABS strategy generally shows good biocompatibility, high selectivity, excellent reliability and high signal contrast, strongly indicating its great potential for high-tech innovations in the sensing field, among which bioprobing is of particular interest. With this Account, we hope to spark new ideas and inspire new endeavors in this emerging research area, further promoting state-of-the-art developments in the field of sensing.