Showing papers by "Panasonic published in 2020"

Single Nanometer-Sized NiFe-Layered Double Hydroxides as Anode Catalyst in Anion Exchange Membrane Water Electrolysis Cell with Energy Conversion Efficiency of 74.7% at 1.0 A cm–2

Abstract: We used a liquid phase reaction to synthesize a nickel (Ni) and iron (Fe)-containing layered double hydroxide (NiFe-LDH), having a lateral size less than 10 nm. A chelating agent introduced into th...

Compressed Volumetric Heatmaps for Multi-Person 3D Pose Estimation

Abstract: In this paper we present a novel approach for bottom-up multi-person 3D human pose estimation from monocular RGB images. We propose to use high resolution volumetric heatmaps to model joint locations, devising a simple and effective compression method to drastically reduce the size of this representation. At the core of the proposed method lies our Volumetric Heatmap Autoencoder, a fully-convolutional network tasked with the compression of ground-truth heatmaps into a dense intermediate representation. A second model, the Code Predictor, is then trained to predict these codes, which can be decompressed at test time to re-obtain the original representation. Our experimental evaluation shows that our method performs favorably when compared to state of the art on both multi-person and single-person 3D human pose estimation datasets and, thanks to our novel compression strategy, can process full-HD images at the constant runtime of 8 fps regardless of the number of subjects in the scene. Code and models are publicly available.

General Video Coding Technology in Responses to the Joint Call for Proposals on Video Compression With Capability Beyond HEVC

Abstract: After the development of the High-Efficiency Video Coding Standard (HEVC), ITU-T VCEG and ISO/IEC MPEG formed the Joint Video Exploration Team (JVET), which started exploring video coding technology with higher coding efficiency, including development of a Joint Exploration Model (JEM) algorithm and a corresponding software implementation. The technology explored in the last version of the JEM further increases the compression capabilities of the hybrid video coding approach by adding new tools, reaching up to 30% bit rate reduction compared to HEVC based on the Bjontegaard delta bit rate (BD-rate) metric, and further improvement beyond that in terms of subjective visual quality. This provided enough evidence to issue a joint Call for Proposals (CfP) for a new standardization activity now known as Versatile Video Coding (VVC). All technology proposed in the responses to the CfP was based on the classic block-based hybrid video coding design, extending it by new elements of partitioning, intra- and inter-picture prediction, prediction signal filtering, transforms, quantization/scaling, entropy coding, and in-loop filtering. This article provides an overview of technology that was proposed in the responses to the CfP, with a focus on techniques that were not already explored in the JEM context.

Silica gel solid nanocomposite electrolytes with interfacial conductivity promotion exceeding the bulk Li-ion conductivity of the ionic liquid electrolyte filler

Abstract: The transition to solid-state Li-ion batteries will enable progress toward energy densities of 1000 W·hour/liter and beyond. Composites of a mesoporous oxide matrix filled with nonvolatile ionic liquid electrolyte fillers have been explored as a solid electrolyte option. However, the simple confinement of electrolyte solutions inside nanometer-sized pores leads to lower ion conductivity as viscosity increases. Here, we demonstrate that the Li-ion conductivity of nanocomposites consisting of a mesoporous silica monolith with an ionic liquid electrolyte filler can be several times higher than that of the pure ionic liquid electrolyte through the introduction of an interfacial ice layer. Strong adsorption and ordering of the ionic liquid molecules render them immobile and solid-like as for the interfacial ice layer itself. The dipole over the adsorbate mesophase layer results in solvation of the Li+ ions for enhanced conduction. The demonstrated principle of ion conduction enhancement can be applied to different ion systems.

Comprehensive Drowsiness Level Detection Model Combining Multimodal Information

Abstract: This paper presents a drowsiness detection model that is capable of sensing the entire range of stages of drowsiness, from weak to strong. The key assumption underlying our approach is that the sitting posture-related index can indicate weak drowsiness that drivers themselves do not notice. We first determined the sensitivity of the posture index and conventional indices for the stages of drowsiness. Then, we designed a drowsiness detection model combining several indices sensitive to weak drowsiness and to strong drowsiness, to cover all drowsiness stages. Subsequently, the model was trained and evaluated on a dataset comprised of data collected from approximately 50 drivers in simulated driving experiments. The results indicated that posture information improved the accuracy of weak drowsiness detection, and our proposed model using the driver’s blink and posture information covered all stages of drowsiness (F1-score 53.6%, root mean square error 0.620). Future applications of this model include not only warning systems for dangerously drowsy drivers but also systems which can take action before their drivers become drowsy. Since measuring the information requires no restrictive equipment such as on-body electrodes, the model presented here based on blink and posture information can be used in several practical applications.

Dreaming: Model-based Reinforcement Learning by Latent Imagination without Reconstruction

Abstract: In the present paper, we propose a decoder-free extension of Dreamer, a leading model-based reinforcement learning (MBRL) method from pixels. Dreamer is a sample- and cost-efficient solution to robot learning, as it is used to train latent state-space models based on a variational autoencoder and to conduct policy optimization by latent trajectory imagination. However, this autoencoding based approach often causes object vanishing, in which the autoencoder fails to perceives key objects for solving control tasks, and thus significantly limiting Dreamer's potential. This work aims to relieve this Dreamer's bottleneck and enhance its performance by means of removing the decoder. For this purpose, we firstly derive a likelihood-free and InfoMax objective of contrastive learning from the evidence lower bound of Dreamer. Secondly, we incorporate two components, (i) independent linear dynamics and (ii) the random crop data augmentation, to the learning scheme so as to improve the training performance. In comparison to Dreamer and other recent model-free reinforcement learning methods, our newly devised Dreamer with InfoMax and without generative decoder (Dreaming) achieves the best scores on 5 difficult simulated robotics tasks, in which Dreamer suffers from object vanishing.

Dynamic imaging of lithium in solid-state batteries by operando electron energy-loss spectroscopy with sparse coding

Abstract: Lithium-ion transport in cathodes, anodes, solid electrolytes, and through their interfaces plays a crucial role in the electrochemical performance of solid-state lithium-ion batteries. Direct visualization of the lithium-ion dynamics at the nanoscale provides valuable insight for understanding the fundamental ion behaviour in batteries. Here, we report the dynamic changes of lithium-ion movement in a solid-state battery under charge and discharge reactions by time-resolved operando electron energy-loss spectroscopy with scanning transmission electron microscopy. Applying image denoising and super-resolution via sparse coding drastically improves the temporal and spatial resolution of lithium imaging. Dynamic observation reveals that the lithium ions in the lithium cobaltite cathode are complicatedly extracted with diffusion through the lithium cobaltite domain boundaries during charging. Even in the open-circuit state, they move inside the cathode. Operando electron energy-loss spectroscopy with sparse coding is a promising combination to visualize the ion dynamics and clarify the fundamentals of solid-state electrochemistry.

Base station, terminal, and communication method

Abstract: A signal assignment unit (105) assigns a downlink control signal including resource assignment information of a PDSCH to a downlink resource. A specification unit (108) specifies a PUCCH resource using an offset value set to either a first PRB set or a second PRB set when the downlink control signal is disposed to spread over the first PRB set and the second PRB set. A signal separation unit (109) separates an ACK/NACK signal included in the specified PUCCH resource from a received signal from a terminal to which the downlink control signal has been transmitted.

System for augmented human–robot interaction through mixed reality and robot training by non-experts in customer service environments

Abstract: Human–robot interaction during general service tasks in home or retail environment has been proven challenging, partly because (1) robots lack high-level context-based cognition and (2) humans cann...

Minimizing the Trade-Off between Photocurrent and Photovoltage in Triple-Cation Mixed-Halide Perovskite Solar Cells

Abstract: Its lower bandgap makes formamidinium lead iodide (FAPbI3) a more suitable candidate for single-junction solar cells than pure methylammonium lead iodide (MAPbI3). However, its structural and thermodynamic stability is improved by introducing a significant amount of MA and bromide, both of which increase the bandgap and amplify trade-off between the photocurrent and photovoltage. Here, we simultaneously stabilized FAPbI3 into a cubic lattice and minimized the formation of photoinactive phases such as hexagonal FAPbI3 and PbI2 by introducing 5% MAPbBr3, as revealed by synchrotron X-ray scattering. We were able to stabilize the composition (FA0.95MA0.05Cs0.05)Pb(I0.95Br0.05)3, which exhibits a minimal trade-off between the photocurrent and photovoltage. This material shows low energetic disorder and improved charge-carrier dynamics as revealed by photothermal deflection spectroscopy (PDS) and transient absorption spectroscopy (TAS), respectively. This allowed the fabrication of operationally stable perovskite solar cells yielding reproducible efficiencies approaching 22%.

PlaNet of the Bayesians: Reconsidering and Improving Deep Planning Network by Incorporating Bayesian Inference

Abstract: In the present paper, we propose an extension of the Deep Planning Network (PlaNet), also referred to as PlaNet of the Bayesians (PlaNet-Bayes). There has been a growing demand in model predictive control (MPC) in partially observable environments in which complete information is unavailable because of, for example, lack of expensive sensors. PlaNet is a promising solution to realize such latent MPC, as it is used to train state-space models via model-based reinforcement learning (MBRL) and to conduct planning in the latent space. However, recent state-of-the-art strategies mentioned in MBRR literature, such as involving uncertainty into training and planning, have not been considered, significantly suppressing the training performance. The proposed extension is to make PlaNet uncertainty-aware on the basis of Bayesian inference, in which both model and action uncertainty are incorporated. Uncertainty in latent models is represented using a neural network ensemble to approximately infer model posteriors. The ensemble of optimal action candidates is also employed to capture multimodal uncertainty in the optimality. The concept of the action ensemble relies on a general variational inference MPC (VI-MPC) framework and its instance, probabilistic action ensemble with trajectory sampling (PaETS). In this paper, we extend VI-MPC and PaETS, which have been originally introduced in previous literature, to address partially observable cases. We experimentally compare the performances on continuous control tasks, and conclude that our method can consistently improve the asymptotic performance compared with PlaNet.

Research Project to Realize Various High-reliability Communications in Advanced 5G Network

Abstract: We started a new research project in the “advanced 5G” era that aims at accommodating various types of communications involving current and emerging services with different data flow-level quality requirements. In this paper, the objectives and the technical aspects of the research project are introduced. We propose an architecture based on a virtualized radio access network (vRAN) that enables adaptive control of equipment resources and location of functions in the vRAN environment in accordance with spatially and temporally changing communication demands. The seven planned research items that are essential for realizing the advanced 5G network are listed as follows: blockage prediction, new radio access technologies (RATs) and their implementations with software-defined radio (SDR), adaptive interference and resource control, integration of radio and fiber resource control, highly efficient access transmission control, adaptive placement of BS functions, and quality aware traffic pattern prediction.

Monolithic Bidirectional Switch Based on GaN Gate Injection Transistors

Abstract: The paper deals with a bi-directional switch based on N-channel enhancement-mode GaN FET. The proposed device is a Gate Injection Transistor monolithic solution to reduce the volume of the switch with high current density and blocking voltage of 600V. It features a dual-gate control pin and two power terminal. In the paper, the main characteristics of the bi-directional switch and the performance in the four-quadrant of operation are examined and discussed. The device characteristics are compared with the traditional MOSFET and IGBT solutions. The gate driver design issues are considered to optimize the switching transient of the GaN-based switch. Finally, an experimental evaluation of the GaN FET as the bidirectional circuit breaker is carried out in an AC power supply system to validate the effectiveness of the proposed monolithic new device.

Encoder, decoder, encoding method, and decoding method

Abstract: An encoder which encodes image information includes memory and circuitry accessible to the memory. The circuitry binarizes a data value indicating the number of non-zero coefficients included in a current basic block which is one of one or more basic blocks in a frequency transform block, according to a conversion table, to encode the image information which includes the data value. When binarizing the data value, the circuitry selects the conversion table from a plurality of tables including two or more tables which differ from each other in difference between a longest bit length and a shortest bit length of a plurality of binary values associated with a plurality of data values, according to the position of the current basic block in the current frequency transform block which is the frequency transform block including the current basic block, and binarizes the data value according to the conversion table selected.

Long-term heat-storage ceramics absorbing thermal energy from hot water.

Abstract: In thermal and nuclear power plants, 70% of the generated thermal energy is lost as waste heat The temperature of the waste heat is below the boiling temperature of water Here, we show a long-term heat-storage material that absorbs heat energy at warm temperatures from 38°C (311 K) to 67°C (340 K) This unique series of material is composed of scandium-substituted lambda-trititanium-pentoxide (λ-Sc x Ti3-x O5) λ-Sc x Ti3-x O5 not only accumulates heat energy from hot water but also could release the accumulated heat energy by the application of pressure λ-Sc x Ti3-x O5 has the potential to accumulate heat energy of hot water generated in thermal and nuclear power plants and to recycle the accumulated heat energy on demand by applying external pressure Furthermore, it may be used to recycle waste heat in industrial factories and automobiles

Improving the Open-Circuit Voltage of Sn-Based Perovskite Solar Cells by Band Alignment at the Electron Transport Layer/Perovskite Layer Interface.

Abstract: Organic–inorganic lead halide perovskites are promising materials for realization of low-cost and high-efficiency solar cells. Because of the toxicity of lead, Sn-based perovskite materials have be...

Visualization of Lithium Transfer Resistance in Secondary Particle Cathodes of Bulk-Type Solid-State Batteries

Abstract: The rate capability of bulk-type solid-state Li-ion batteries is limited by the solid-state diffusion of Li ions in the cathodes that typically form micrometer-sized polycrystalline particles compo...

Machine-Learning Guided Quantum Chemical and Molecular Dynamics Calculations to Design Novel Hole-Conducting Organic Materials

Abstract: Materials exhibiting higher mobilities than conventional organic semiconducting materials such as fullerenes and fused thiophenes are in high demand for applications such as printed electronics, organic solar cells, and image sensors. In order to discover new molecules that might show improved charge mobility, combined density functional theory (DFT) and molecular dynamics (MD) calculations were performed, guided by predictions from machine learning (ML). A ML model was constructed based on 32 values of theoretically calculated hole mobilities for thiophene derivatives, benzodifuran derivatives, a carbazole derivative and a perylene diimide derivative with the maximum value of 10-1.96 cm2/(V s). Sequential learning, also known as active learning, was applied to select compounds on which to perform DFT/MD calculation of hole mobility to simultaneously improve the mobility surrogate model and identify high mobility compounds. By performing 60 cycles of sequential learning with 165 DFT/MD calculations, a molecule having a fused thioacene structure with its calculated hole mobility of 10-1.86 cm2/(V s) was identified. This values is higher than the maximum value of mobility in the initial training data set, showing that an extrapolative discovery could be made with the sequential learning.

Restock and straightening system for retail automation using compliant and mobile manipulation

Abstract: As the retail industry keeps expanding and shortage of workers increasing, there is a need for autonomous manipulation of products to support retail operations. The increasing amount of products an...

Deep Active Learning for Biased Datasets via Fisher Kernel Self-Supervision

Abstract: Active learning (AL) aims to minimize labeling efforts for data-demanding deep neural networks (DNNs) by selecting the most representative data points for annotation. However, currently used methods are ill-equipped to deal with biased data. The main motivation of this paper is to consider a realistic setting for pool-based semi-supervised AL, where the unlabeled collection of train data is biased. We theoretically derive an optimal acquisition function for AL in this setting. It can be formulated as distribution shift minimization between unlabeled train data and weakly-labeled validation dataset. To implement such acquisition function, we propose a low-complexity method for feature density matching using self-supervised Fisher kernel (FK) as well as several novel pseudo-label estimators. Our FK-based method outperforms state-of-the-art methods on MNIST, SVHN, and ImageNet classification while requiring only 1/10th of processing. The conducted experiments show at least 40% drop in labeling efforts for the biased class-imbalanced data compared to existing methods.

Annotating and Extracting Synthesis Process of All-Solid-State Batteries from Scientific Literature

Abstract: The synthesis process is essential for achieving computational experiment design in the field of inorganic materials chemistry. In this work, we present a novel corpus of the synthesis process for all-solid-state batteries and an automated machine reading system for extracting the synthesis processes buried in the scientific literature. We define the representation of the synthesis processes using flow graphs, and create a corpus from the experimental sections of 243 papers. The automated machine-reading system is developed by a deep learning-based sequence tagger and simple heuristic rule-based relation extractor. Our experimental results demonstrate that the sequence tagger with the optimal setting can detect the entities with a macro-averaged F1 score of 0.826, while the rule-based relation extractor can achieve high performance with a macro-averaged F1 score of 0.887.

Cyclopentadithiophene-Based Hole-Transporting Material for Highly Stable Perovskite Solar Cells with Stabilized Efficiencies Approaching 21%

Abstract: There is an urge to develop new hole-transporting materials (HTMs) for perovskite solar cells (PSCs), which can yield comparable power conversion efficiencies (PCEs) yet mitigate the issue of stability associated with the state-of-the-art HTM Spiro-MeOTAD Herein, we designed and prepared C-2v-symmetric spiro-configured HTM-1 comprising a central acridine-cyclopentadithiophene core unit flanked with triarylamine moieties PSCs containing a 40 nm thin HTM-1 layer for hole extraction yielded a stabilized PCE approaching 21% under standard illumination Owing to its higher hole mobility (mu(h)) at low electric field, an impressive short-circuit current density (J(SC)) of 247 mA cm(-2) and a high fill factor (FF) of 077 have been achieved More importantly, HTM-1-based PSCs presented an excellent long-term operational stability under continuous illumination for 400 h and thermal stability at 80 degrees C, which can be ascribed to its high glass transition temperature of 168 degrees C and superior moisture tolerance Arguably, the confluence of high performance and remarkable stability will lead to the development of technologically interesting new, stable, and efficient PSCs

Automatic Detection of Cognitive Impairments through Acoustic Analysis of Speech.

Abstract: Background Early detection of mild cognitive impairment is crucial in the prevention of Alzheimer's disease. The aim of the present study was to identify whether acoustic features can help differentiate older, independent community-dwelling individuals with cognitive impairment from healthy controls. Methods A total of 8779 participants (mean age 74.2 ± 5.7 in the range of 65-96, 3907 males and 4872 females) with different cognitive profiles, namely healthy controls, mild cognitive impairment, global cognitive impairment (defined as a Mini Mental State Examination score of 20-23), and mild cognitive impairment with global cognitive impairment (a combined status of mild cognitive impairment and global cognitive impairment), were evaluated in short-sentence reading tasks, and their acoustic features, including temporal features (such as duration of utterance, number and length of pauses) and spectral features (F0, F1, and F2), were used to build a machine learning model to predict their cognitive impairments. Results The classification metrics from the healthy controls were evaluated through the area under the receiver operating characteristic curve and were found to be 0.61, 0.67, and 0.77 for mild cognitive impairment, global cognitive impairment, and mild cognitive impairment with global cognitive impairment, respectively. Conclusion Our machine learning model revealed that individuals' acoustic features can be employed to discriminate between healthy controls and those with mild cognitive impairment with global cognitive impairment, which is a more severe form of cognitive impairment compared with mild cognitive impairment or global cognitive impairment alone. It is suggested that language impairment increases in severity with cognitive impairment.

Ultimate suppression of thermal transport in amorphous silicon nitride by phononic nanostructure

Abstract: Engineering the thermal conductivity of amorphous materials is highly essential for the thermal management of future electronic devices. Here, we demonstrate the impact of ultrafine nanostructuring on the thermal conductivity reduction of amorphous silicon nitride (a-Si3N4) thin films, in which the thermal transport is inherently impeded by the atomic disorders. Ultrafine nanostructuring with feature sizes below 20 nm allows us to fully suppress contribution of the propagating vibrational modes (propagons), leaving only the diffusive vibrational modes (diffusons) to contribute to thermal transport in a-Si3N4. A combination of the phonon-gas kinetics model and the Allen-Feldmann theory reproduced the measured results without any fitting parameters. The thermal conductivity reduction was explained as extremely strong diffusive boundary scattering of both propagons and diffusons. These findings give rise to substantial tunability of thermal conductivity of amorphous materials, which enables us to provide better thermal solutions in microelectronic devices.

High-Brightness Red-Emitting Phosphor La3(Si,Al)6(O,N)11:Ce3+ for Next-Generation Solid-State Light Sources.

Abstract: A novel high-brightness red-emitting phosphor, La3(Si,Al)6(O,N)11:Ce3+ (LSA), which can potentially be used as a laser-excited light source, is demonstrated Laser-excited phosphor system has great potential for use as a white-light source, as it is orders of magnitude brighter than white LEDs Although conventional yellow-green phosphors show excellent luminescent properties even under high-power laser excitation, red-emitting phosphors, which are essential to achieve a high color-rendering index and low color-temperature, show quantum efficiency quenching This limits the output power in multiphosphor excitation systems Ce3+ should successfully tolerate high-power excitation due to the shortest emission lifetime seen in rare-earth ions, caused by the 5d1-4f1 spin-allowed transition; however, a red-emitting Ce3+-doped phosphor of practical use has not been realized LSA is described by the crystal-field modification of a yellow-emitting phosphor, La3Si6N11:Ce3+, with substitution of Al in Si sites LSA shows 640 nm red emission together with tolerance for high-power excitation and thermal quenching, suggesting its significant potential for industrial applications that require ultrahigh brightness

Adaptive motion generation using imitation learning and highly compliant end effector for autonomous cleaning

Abstract: Recent demographic trends in super aging societies, such as Japan, is leading to severe worker shortage. Service robots can play a promising role to augment human workers for performing various hou...

Image encoder, image decoder, image encoding method, and image decoding method

Abstract: An image encoder is provided, which includes circuitry and a memory coupled to the circuitry The circuitry, in operation, performs a boundary smoothing operation along a boundary between a first partition having a non-rectangular shape (eg, a triangular shape) and a second partition that are split from an image block The boundary smoothing operation includes: first-predicting first values of a set of pixels of the first partition along the boundary, using information of the first partition; second-predicting second values of the set of pixels of the first partition along the boundary, using information of the second partition; weighting the first values and the second values; and encoding the first partition using the weighted first values and the weighted second values