A5100322208- Human Pose and Action Recognition
- Spectroscopy and Chemometric Analyses
- Advanced Vision and Imaging
- Hand Gesture Recognition Systems
- Spectroscopy Techniques in Biomedical and Chemical Research
- 3D Shape Modeling and Analysis
- Speech and Audio Processing
- Invertebrate Immune Response Mechanisms
- Anomaly Detection Techniques and Applications
- Music and Audio Processing
- Elasticity and Material Modeling
- Neurobiology and Insect Physiology Research
- Advanced Chemical Sensor Technologies
- Generative Adversarial Networks and Image Synthesis
- Insect symbiosis and bacterial influences
- Robotics and Sensor-Based Localization
- Face recognition and analysis
- Aluminum toxicity and tolerance in plants and animals
- Multimodal Machine Learning Applications
- Smart Agriculture and AI
- Meat and Animal Product Quality
- Wireless Body Area Networks
- Polysaccharides and Plant Cell Walls
- Augmented Reality Applications
- Gait Recognition and Analysis
Clarkson University
2025
The University of Queensland
2023-2024
NetEase (China)
2023-2024
Dalian Maritime University
2024
Shaanxi Railway Institute
2024
First Affiliated Hospital of Henan University
2024
Beijing University of Chemical Technology
2022-2023
Northwest A&F University
2014-2023
Agency for Science, Technology and Research
2023
Institute of Microelectronics
2023
This paper proposes a deep neural network (DNN) for piece-wise planar depthmap reconstruction from single RGB image. While DNNs have brought remarkable progress to single-image depth prediction, requires structured geometry representation, and has been difficult task master even DNNs. The proposed end-to-end DNN learns directly infer set of plane parameters corresponding segmentation masks We generated more than 50,000 depthmaps training testing ScanNet, large-scale RGBD video database. Our...
Generating vivid and diverse 3D co-speech gestures is crucial for various applications in animating virtual avatars. While most existing methods can generate from audio directly, they usually overlook that <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"/> <italic xmlns:xlink="http://www.w3.org/1999/xlink"><b>emotion</b></i> one of the key factors authentic gesture generation. In this work, we propose...
Given an audio-visual pair, segmentation (AVS) aims to locate sounding sources by predicting pixel-wise maps. Previous methods assume that each sound component in audio signal always has a visual counterpart the image. However, this assumption overlooks off-screen sounds and background noise often contaminate recordings real-world scenarios. They impose significant challenges on building consistent semantic mapping between signals for AVS models thus impede precise localization. In work, we...
The audio-visual segmentation (AVS) task aims to segment sounding objects from a given video. Existing works mainly focus on fusing audio and visual features of video achieve object masks. However, we observed that prior arts are prone certain salient in regardless the information. This is because often most ones AVS dataset. Thus, current methods might fail localize genuine due dataset bias. In this work, present an instance-aware approach overcome nutshell, our method first localizes...
Predicting natural and diverse 3D hand gestures from the upper body dynamics is a practical yet challenging task in virtual avatar creation. Previous works usually overlook asymmetric motions between two hands generate holistic manner, leading to unnatural results. In this work, we introduce novel bilateral disentanglement based two-stage generation method achieve prediction dynamics. first stage, intend by hand-disentanglement branches. Considering of hands, Spatial-Residual Memory (SRM)...
Oil body emulsions (OBEs) affect the final oil yield as an intermediate in concurrent peanut and protein extraction process using aqueous enzyme (AEE) method. Roasting temperature promotes cell structure breakdown, affecting OBE composition stability improving rates. Therefore, this study aimed to investigate effects of pretreatment at different roasting temperatures on extracted through AEE. The results showed that rates peaked 90 °C, 92.21%, 77.02%, respectively. did not change but...
For exploring vegetarian eating in contemporary Guangzhou, this article employs the concept of ‘ethical eating’ and framework ethical consumption as outlined by Clive Barnett his coauthors to examine food practices a restaurant through two dimensions: dimension consumer (2005b). Drawing on ethnographic observations interviews conducted Refreshing Vegetarian Food (Yixin Su Shi), capital city Guangdong province, argues that, one hand, can provide an environment for connects ethics human bodies...
A critical aspect of human space exploration and eventual settlement is the ability to construct habitats while minimizing payload mass launched from Earth. To respond this challenge, we have proposed use fungal bio-composites for growing extra-terrestrial structures, directly at destination, significantly lowering structural materials transported Earth need high robotic operations infrastructure preparations. Throughout history, construction has used biologically produced materials, bone...
It is an urgent task to improve the applicability of cucumber disease classification model in greenhouse edge-intelligent devices. The energy consumption diagnosis models designed based on deep learning methods a key factor affecting its applicability. Based this motivation, two reducing model’s calculation amount and changing method feature extraction were used study reduce consumption, thereby prolonging working time edge devices deployed with models. First, dataset complex backgrounds...
While deep learning has been widely used for video analytics, such as classification and action detection, dense detection with fast-moving subjects from sports videos is still challenging. In this work, we release yet another benchmark P 2 ANet ing ong- A ction which consists of 2,721 clips collected the broadcasting professional table tennis matches in World Table Tennis Championships Olympiads. We work a crew professionals referees on specially designed annotation toolbox to obtain...