Hong Liu

• Central South University

Human Mesh Reconstruction (HMR) from monocular video plays an important role in human-robot interaction and collaboration. However, existing video-based human mesh reconstruction methods face a trade-off between accurate reconstruction and smooth motion. These methods design networks based on either RNNs or attention mechanisms to extract local tem...

Considering the instance-level discriminative ability, contrastive learning methods, including MoCo and SimCLR, have been adapted from the original image representation learning task to solve the self-supervised skeleton-based action recognition task. These methods usually use multiple data streams (i.e., joint, motion, and bone) for ensemble learn...

Underwater object detection (UOD) is crucial for marine economic development, environmental protection, and the planet's sustainable development. The main challenges of this task arise from low‐contrast, small objects, and mimicry of aquatic organisms. The key to addressing these challenges is to focus the model on obtaining more discriminative inf...

Occluded person re-identification is a challenging problem due to the destruction of occluders in different camera views. Most existing paradigms focus on visible human body parts through some external models to reduce noise interference. However, the feature misalignment problem caused by discarded occlusions negatively affects the performance of...

Regression-based 3D human pose and shape estimation often fall into one of two different paradigms. Parametric approaches, which regress the parameters of a human body model, tend to produce physically plausible but image-mesh misalignment results. In contrast, non-parametric approaches directly regress human mesh vertices, resulting in pixel-align...

Considering the instance-level discriminative ability, contrastive learning methods, including MoCo and SimCLR, have been adapted from the original image representation learning task to solve the self-supervised skeleton-based action recognition task. These methods usually use multiple data streams (i.e., joint, motion, and bone) for ensemble learn...

Underwater object detection (UOD) plays a significant role in aquaculture and marine environmental protection. Considering the challenges posed by low contrast and low-light conditions in underwater environments, several underwater image enhancement (UIE) methods have been proposed to improve the quality of underwater images. However, only using th...

Underwater object detection (UOD) is crucial for marine economic development, environmental protection, and the planet's sustainable development. The main challenges of this task arise from low-contrast, small objects, and mimicry of aquatic organisms. The key to addressing these challenges is to focus the model on obtaining more discriminative inf...

Robot motion planning is an important part of the unmanned supermarket. The challenges of motion planning in supermarkets lie in the diversity of the supermarket environment, the complexity of obstacle movement, the vastness of the search space. This paper proposes an adaptive Search and Path planning method based on the Semantic information and De...

Recently, deep reinforcement learning (DRL) methods have significantly improved the performance of target-driven indoor navigation tasks. However, the rich semantic information of environments is still not fully exploited in previous approaches. In addition, existing methods usually tend to overfit on training scenes or objects in target-driven nav...

One of the challenging problems in the research field of polymer nanocomposites is how to prepare nanocomposites with high grafting density and strong ability of dispersion at the same time. For nanocomposites composed of bimodal bidisperse polymer chains and nanoparticles, the above requirements can be met by rationally adjusting the ratio of long...

We propose a possible strategy that may experimentally generate long polymeric chains with an entanglement-free structure. The basic idea is designing the conditions to restrict polymer chains from growing along the surface with an obviously concave curvature. This strategy is proved to effectively reduce the chance of forming both inter- and intra...

The self‐assembly of polymer‐grafted nanoparticles is increasingly applied in the field of functional materials. However, the corresponding relationship between the intrinsic dynamic transition path and microstructure is still not clear enough, which will lead to an inability to achieve further precise regulation and directional design in experimen...

By employing dissipative particle dynamics (DPD) simulations combined with stochastic polymerization models, we have conducted a detailed simulation study of supramolecular solution polymerization as well as interfacial polymerization employing a coarse-grained model which is closer to the real monomer structure. By adding bending angle potentials...

In recent years, self-supervised representation learning for skeleton-based action recognition has been developed with the advance of contrastive learning methods. The existing contrastive learning methods use normal augmentations to construct similar positive samples, which limits the ability to explore novel movement patterns. In this paper, to m...

Complicated underwater environments bring new challenges to object detection, such as unbalanced light conditions, low contrast, occlusion, and mimicry of aquatic organisms. Under these circumstances, the objects captured by the underwater camera will become vague, and the generic detectors often fail on these vague objects. This work aims to solve...

Modern multi-layer perceptron (MLP) models have shown competitive results in learning visual representations without self-attention. However, existing MLP models are not good at capturing local details and lack prior knowledge of human configurations, which limits their modeling power for skeletal representation learning. To address these issues, w...

Arbitrary-oriented object detection (AOOD) is a challenging task to detect objects in the wild with arbitrary orientations and cluttered arrangements. Existing approaches are mainly based on anchor-based boxes or dense points, which rely on complicated hand-designed processing steps and inductive bias, such as anchor generation, transformation, and...

To explore the nature of water insolubility of polyformaldehyde (POM), we utilize single-molecule atomic force microscopy (AFM) and molecular dynamics (MD) simulations to study the possible interactions between POM and water at the single-chain level. Single-molecule force-extension curves of POM obtained in nonane and deionized water showed a mark...

Molecules adsorbed or attached on a surface is a quite basic phenomenon in numerous chemical or biological systems. Grafting-onto is considered as a feasible way to achieve it. The grafting reaction is essentially controlled by the diffusion of the molecules, thus it is more likely a physical issue, instead of a chemical issue. Because of the exper...

Arbitrary-oriented object detection (AOOD) is a challenging task to detect objects in the wild with arbitrary orientations and cluttered arrangements. Existing approaches are mainly based on anchor-based boxes or dense points, which rely on complicated hand-designed processing steps and inductive bias, such as anchor generation, transformation, and...

Although monocular 3D human pose estimation methods have made significant progress, it's far from being solved due to the inherent depth ambiguity. Instead, exploiting multi-view information is a practical way to achieve absolute 3D human pose estimation. In this paper, we propose a simple yet effective pipeline for weakly-supervised cross-view 3D...

In recent years, self-supervised representation learning for skeleton-based action recognition has been developed with the advance of contrastive learning methods. The existing contrastive learning methods use normal augmentations to construct similar positive samples, which limits the ability to explore novel movement patterns. In this paper, to m...

In the last decade, polymerization‐induced self‐assembly (PISA) has emerged to be a quite popular technique for preparing a variety of nanoassembly structures in a selective solvent. As a special technique of PISA, polymerization‐induced cooperative assembly (PICA) is believed to be able to produce more variable nanostructures and its ways of regul...

Although monocular 3D human pose estimation methods have made significant progress, it's far from being solved due to the inherent depth ambiguity. Instead, exploiting multi-view information is a practical way to achieve absolute 3D human pose estimation. In this paper, we propose a simple yet effective pipeline for weakly-supervised cross-view 3D...

The composition and structure of a membrane determine its functionality and practical application. We study the supramolecular polymeric membrane prepared by supramolecular emulsion interfacial polymerization (SEIP) on the oil-in-water droplet via the computer simulation method. The factors that may influence its structure and properties are invest...

A multiscale simulation strategy was proposed to study the curing reaction on the network formation and corresponding mechanical properties of a bio-based epoxy resin composite. The crosslinking process of the system to form an epoxy network structure was reproduced on the mesoscopic scale by the dissipative particle dynamics simulation coupled wit...

The performance of existing underwater object detection methods degrades seriously when facing domain shift problem caused by complicated underwater environments. Due to the limitation of the number of domains in the dataset, deep detectors easily just memorize a few seen domain, which leads to low generalization ability. Ulteriorly, it can be infe...

Despite great progress in video-based 3D human pose estimation, it is still challenging to learn a discriminative single-pose representation from redundant sequences. To this end, we propose a novel Transformer-based architecture, called Lifting Transformer, for 3D human pose estimation to lift a sequence of 2D joint locations to a 3D pose. Specifi...

Synthesizing polymers with tailor-made molecular weight distribution (MWD) is an essential step toward better control and design of functional polymer materials. We propose a novel one-pot reaction strategy that can facilitate the inverse design of the shape, breadth, and skew of the MWD in a controlled polymerization. By a multistep initiator addi...

Because of the important application in nanocomposite materials, the preparation of nanoparticles (NPs) tethered with bimodal bidispersed polymer chains via grafting-from strategy is studied by coarse-grained molecular dynamics simulations combined with the stochastic reaction model. We design three conventional scenarios for preparing these NPs, i...

A supramolecular diblock copolymer formed by reversible bonds between the two blocks shows a rich microphase separation behavior and has great application potential in stimuli-responsive materials. We propose a novel method to describe supramolecular reactions in dissipative particle dynamics, which includes a reversible reaction to accurately repr...

3D skeleton-based action recognition, owing to the latent advantages of skeleton, has been an active topic in computer vision. As a consequence, there are lots of impressive works including conventional handcraft feature based and learned feature based have been done over the years. However, previous surveys about action recognition mostly focus on...

The ligand shell of a nanoparticle (NP) determines most of the interfacial properties through its composition and structure. Despite widespread study over the years, the factors impacting the ligand shell structures, especially the effects of ligand adsorption kinetics in solution, are still not clear and even conflict with each other. We have deve...

Tiny defect detection (TDD) which aims to perform the quality control of printed circuit boards (PCBs) is a basic and essential task in the production of most electronic products. Though significant progress has been made in PCB defect detection, traditional methods are still difficult to cope with the complex and diverse PCBs. To deal with these p...

The vitrimer with dynamic covalent bond makes the thermosetting material plastic, recyclable and self-repairing, and has broad application prospects. However, due to the complex composition of vitrimer and the dynamic bond exchange reactions (BERs), the mechanism behind its unique dynamic behavior is not fully understood. We used the hybrid molecul...

This paper presents a new framework for human action recognition by fusing human motion with skeletal joints. Firstly, Adaptive Hierarchical Depth Motion Maps (AH-DMMs) are proposed to capture shape and motion cues of action sequences. Specifically, AH-DMMs are calculated over adaptive hierarchical windows and Gabor filters are used to encode the t...

Skeleton-based human action recognition (HAR) has attracted a lot of research attentions because of robustness to variations of locations and appearances. However, most existing methods treat the whole skeleton as a fixed pattern, in which the importance of different skeleton joints for action recognition is not considered. In this paper, a novel C...

Rain removal aims to extract and remove rain streaks from images. Although convolutional neural network (CNN) based methods have achieved impressive results in this field, they are not equivariant to object rotation, which decreases their generalization capacity for tilted rain streaks. In order to solve this problem, we propose Deep Symmetry Enhan...

Keyword1 spotting (KWS) deals with the identification of keywords in speech utterances. A two-stage approach is often used for the flexibility and high efficiency. The two stages are keyword hypotheses detection stage and hit or false-alarm verification stage in sequence. How to reduce the false-alarms is a key and difficult problem in the verifica...

In this study, using the disspative particle dynamics simulations coupled with the stochaistic reaction model, we investigate the polymerization-induced polymer aggregation process and the polymer aggregation-enhanced polymerization process in a binary solution, by simply tuning the solubility of the solvent to one species of copolymerizaiton. Our...

Thioredoxin reductase (TrxR), an antioxidant enzyme dependent on nicotinamide adenine dinucleotide phosphate, plays a vital role in defense against oxidative stress. However, the role of microRNAs targeting TrxR under oxidative stress has not yet been determined. In this study, we tested the involvement of miRNA-mediated posttranscriptional regulat...

Figure S: The effect of pmiR-125a on mature miR-125a expression and the verified target protein P53.

By Brownian dynamics simulations we study the simultaneous polymer chain growth process within the coexistence of bulk and surface initiators. We find when the surface initiator density is low enough, the practical experimental way to estimate the dispersity D of surface-initiated chains on the basis of the dispersity of bulk-initiated chains remai...

The nanoparticles (NPs) grafted with polymer chains prepared by grafting-from strategy is studied by coarse-grained molecular dynamics simulations combined with our stochastic reaction model. A system involving multiple individual NPs with grafting-from processes for all the NPs induced simultaneously is simulated, so that the competitions of chain...

We describe the algorithm of employing multi-GPU power on the basis of Message Passing Interface (MPI) domain decomposition in a molecular dynamics code, GALAMOST, which is designed for the coarse-grained simulation of soft matters. The code of multi-GPU version is developed based on our previous single-GPU version. In multi-GPU runs, one GPU takes...

The development of highly selective, chemically stable and moisture resistant adsorbents is a key milestone for gas separation. Porous carbons featured with random orientation and cross-linking of turbostratic nanodomains usually have wide distribution of micropore. Here we have developed a thermoregulated phase transition assisted synthesis of car...

The development of highly selective, chemically stable and moisture resistant adsorbents is a key milestone for gas separation. Porous carbons featured with random orientation and cross-linking of turbostratic nanodomains usually have wide distribution of micropore. Here we have developed a thermoregulated phase transition assisted synthesis of car...

By constructing a grafting-to reaction model of polydispersed polymer chains to bind on the nanoprticles (NPs), we elucidate the changes of grafting density, poydispersity index and chain length distribution of grafted ligand chains as a dependence of the feeding polymer chains. Our study shows a linear dependence of the grafting density on the ave...

3D action recognition has broad applications in human-computer interaction and intelligent surveillance. However , recognizing similar actions remains challenging since previous literature fails to capture motion and shape cues effectively from noisy depth data. In this paper, we propose a novel two-layer Bag-of-Visual-Words (BoVW) model, which sup...

Human action recognition remains challenging in realistic videos, where scale and viewpoint changes make the problem complicated. Many complex models have been developed to overcome these difficulties, while we explore using low-level features and typical classifiers to achieve the state-of-the-art performance. The baseline model of feature encodin...

We have developed a multiscale model that combines first-principles methods, with atomistic and mesoscopic simulations to explore the molecular structures and packing density of the ligands present on the gold nanoparticles (AuNPs) surface, as well as the adsorption/exchange reaction kinetics of cetyltrimethylammonium bromide (CTAB)/PEG-SH ligands...

This paper presents a local spatio-temporal descriptor for action recognition from depth video sequences which is capable of distinguishing similar actions as well as coping with different speeds of actions. This descriptor is based on three processing stages. In the first stage, the shape and motion cues are captured from a weighted depth sequence...

By molecular dynamics simulations coupled with the stochastic polymerization model, we show a certain dependence of the polymer chain polydispersity on the heterogeneity of the initiator sites distribution during surface-initiated polymerization. The heterogeneity of the initiators will lead to change of the mass distribution and the uncontrollabil...

Action recognition using depth sequences plays important role in many fields, e.g., intelligent surveillance, content-based video retrieval. Real applications require robust and accurate action recognition method. In this paper, we propose a skeleton visualization method, which efficiently encodes the spatial-temporal information of skeleton joints...

It remains a challenge to efficiently represent spatial-temporal data for 3D action recognition. To solve this problem, this paper presents a new skeleton-based action representation using data visualization and convolutional neural networks, which contains four main stages. First, skeletons from an action sequence are mapped as a set of five dimen...

This paper presents an effective local spatial-temporal de-scriptor for action recognition from skeleton sequences. The unique property of our descriptor is that it takes the spatial-temporal discrimination and action speed variations into account , intending to solve the problems of distinguishing similar actions and identifying actions with diffe...

Human action classification, which is vital for content-based video retrieval and human-machine interaction, finds problem in distinguishing similar actions. Previous works typically detect spatial-temporal interest points (STIPs) from action sequences and then adopt bag-of-visual words (BoVW) model to describe actions as numerical statistics of ST...

It remains a challenge to efficiently extract spatialtemporal data from skeleton sequences for 3D human action recognition. Since 3D convolutional neural network(3DCNN) is a powerful tool to simultaneously learn features from both spatial and temporal dimensions, in this paper, we propose a two-stream model using 3DCNN. To our best knowledge, this...

Person re-identification (re-id) on robot platform is an important application for human-robot-interaction (HRI), which aims at making the robot recognize the around persons in varying scenes. Although many effective methods have been proposed for surveillance re-id in recent years, re-id on robot platform is still a novel unsolved problem. Most ex...

Depth motion maps (DMMs) have shown effectiveness in human action recognition, however, they lose the temporal information and suffer from intra-class variations caused by action speed variations. To address these challenges, we propose a novel method for human action recognition. Firstly, Adaptive Hierarchical Depth Motion Maps (AH-DMMs) are calcu...

Human action recognition based on skeletons has wide applications in human-computer interaction and intelligent surveillance. However, view variations and noisy data bring challenges to this task. What's more, it remains a problem to effectively represent spatio-temporal skeleton sequences. To solve these problems in one goal, this work presents an...

This paper presents an effective multi-scale energy-based Global Ternary Image (GTI) representation for action recognition from depth sequences. The unique property of our representation is that it takes the spatial-temporal discrimination and action speed variations into account, intending to solve the problems of distinguishing similar actions an...

To better understand the expression of human smile, there have been considerable studies about automatic smile detection. Despite all the research, few attention is paid to analyse a smile in a comprehensive way. In this paper, a smile analysis system is presented to detailedly measure a person's smile, which consists of three main modules: smile d...

In this work, a molecular dynamics simulation method was introduced to compute the pre-assembled system of molecular imprinted polymers for sulfamethoxazole monomer. The results revealed that the ratio of sulfamethoxazole as template molecule to 3-aminopropyltriethoxysilane as functional monomer to tetraethylorthosilicate as cross-linker of molar r...

Fall detection plays a significant role in the medical care of elderly people. As reported, falling is the main cause leading to the death of the elderly over the age of 65. Recently, video-based fall detection methods have been widely used, with the progress of intelligent video analysis. In this study, we presents a novel method to detect the fal...

Indoor multi-tracking is more challenging compared with outdoor tasks due to frequent occlusion, view-truncation, severe scale change and pose variation, which may bring considerable unreliability and ambiguity to target representation and data association. So discriminative and reliable target representation is vital for accurate data association...

In order to efficiently recognize actions from depth sequences, we propose a novel feature, called Global Ternary Image (GTI), which implicitly encodes both motion regions and motion directions between consecutive depth frames via recording the changes of depth pixels. In this study, each pixel in GTI indicates one of the three possible states, nam...

Recently, approaches utilizing spatial-temporal features to form Bag-of-Words (BoWs) models have achieved great success due to their simplicity and effectiveness. But they still have difficulties when distinguishing between actions with high inter-ambiguity. The main reason is that they describe actions by orderless bag of features, and ignore the...

Anomaly detection is still a challenging task for video surveillance due to complex environments and unpredictable human behaviors. Most existing approaches train offline detectors using manually labeled data and predefined parameters, and are hard to model changing scenes. This paper introduces a neural network based model called online Growing Ne...

Person re-identification (re-id) consists of associating individual across camera network, which is valuable for intelligent video surveillance and has drawn wide attention. Although person re-identification research is making progress, it still faces some challenges such as varying poses, illumination and viewpoints. For feature representation in...

Keyword spotting remainsa challenge when applied to real-world environments with dramatically changing noise. In recent studies, audio-visual integration methods have demonstrated superiorities since visual speech is not influenced by acoustic noise. However, for visual speech recognition, individual utterance mannerisms can lead to confusion and f...