Human Action Recognition Via Multi-modality Information

Gao, Zan,Song, Jian-Ming,Zhang, Hua,Liu, An-An,Xue, Yan-Bing,Xu, Guang-Ping The Korean Institute of Electrical Engineers 2014 Journal of Electrical Engineering & Technology Vol.9 No.2

In this paper, we propose pyramid appearance and global structure action descriptors on both RGB and depth motion history images and a model-free method for human action recognition. In proposed algorithm, we firstly construct motion history image for both RGB and depth channels, at the same time, depth information is employed to filter RGB information, after that, different action descriptors are extracted from depth and RGB MHIs to represent these actions, and then multimodality information collaborative representation and recognition model, in which multi-modality information are put into object function naturally, and information fusion and action recognition also be done together, is proposed to classify human actions. To demonstrate the superiority of the proposed method, we evaluate it on MSR Action3D and DHA datasets, the well-known dataset for human action recognition. Large scale experiment shows our descriptors are robust, stable and efficient, when comparing with the-state-of-the-art algorithms, the performances of our descriptors are better than that of them, further, the performance of combined descriptors is much better than just using sole descriptor. What is more, our proposed model outperforms the state-of-the-art methods on both MSR Action3D and DHA datasets.

Human Action Recognition Via Multi-modality Information

Zan Gao,Jian-ming Song,Hua Zhang,An-An Liu,Yan-bing Xue,Guang-ping Xu 대한전기학회 2014 Journal of Electrical Engineering & Technology Vol.9 No.2

In this paper, we propose pyramid appearance and global structure action descriptors on both RGB and depth motion history images and a model-free method for human action recognition. In proposed algorithm, we firstly construct motion history image for both RGB and depth channels, at the same time, depth information is employed to filter RGB information, after that, different action descriptors are extracted from depth and RGB MHIs to represent these actions, and then multimodality information collaborative representation and recognition model, in which multi-modality information are put into object function naturally, and information fusion and action recognition also be done together, is proposed to classify human actions. To demonstrate the superiority of the proposed method, we evaluate it on MSR Action3D and DHA datasets, the well-known dataset for human action recognition. Large scale experiment shows our descriptors are robust, stable and efficient, when comparing with the-state-of-the-art algorithms, the performances of our descriptors are better than that of them, further, the performance of combined descriptors is much better than just using sole descriptor. What is more, our proposed model outperforms the state-of-the-art methods on both MSR Action3D and DHA datasets.

Human Action Recognition Using Pyramid Histograms of Oriented Gradients and Collaborative Multi-task Learning

( Zan Gao ),( Hua Zhang ),( An-an Liu ),( Yan-bing Xue ),( Guang-ping Xu ) 한국인터넷정보학회 2014 KSII Transactions on Internet and Information Syst Vol.8 No.2

In this paper, human action recognition using pyramid histograms of oriented gradients and collaborative multi-task learning is proposed. First, we accumulate global activities and construct motion history image (MHI) for both RGB and depth channels respectively to encode the dynamics of one action in different modalities, and then different action descriptors are extracted from depth and RGB MHI to represent global textual and structural characteristics of these actions. Specially, average value in hierarchical block, GIST and pyramid histograms of oriented gradients descriptors are employed to represent human motion. To demonstrate the superiority of the proposed method, we evaluate them by KNN, SVM with linear and RBF kernels, SRC and CRC models on DHA dataset, the well-known dataset for human action recognition. Large scale experimental results show our descriptors are robust, stable and efficient, and outperform the state-of-the-art methods. In addition, we investigate the performance of our descriptors further by combining these descriptors on DHA dataset, and observe that the performances of combined descriptors are much better than just using only sole descriptor. With multimodal features, we also propose a collaborative multi-task learning method for model learning and inference based on transfer learning theory. The main contributions lie in four aspects: 1) the proposed encoding the scheme can filter the stationary part of human body and reduce noise interference; 2) different kind of features and models are assessed, and the neighbor gradients information and pyramid layers are very helpful for representing these actions; 3) The proposed model can fuse the features from different modalities regardless of the sensor types, the ranges of the value, and the dimensions of different features; 4) The latent common knowledge among different modalities can be discovered by transfer learning to boost the performance.

FEM Simulation and Test Verification of PD Ultrasonic Signal Propagation in a Power Transformer Model

Zan Wang,Chaofei Gao,Liwei Zheng,Jikun Ren,Wei Wang,Yushuai Zhang,Shijie Han 대한전기학회 2021 Journal of Electrical Engineering & Technology Vol.16 No.1

Ultrasonic signals will be generated when partial discharge occurs in internal insulation faults in large oil immersed power transformers: because the ultrasonic signal has strong anti-interference ability and has no direct electromagnetic contact with the equipment, it is widely used in transformer fault detection and positioning. In this paper, the fi nite element method (FEM) is used to simulate the ultrasonic signal in a 35 kV power transformer. The infl uence of transformer case on ultrasonic signal propagation is considered, and the propagation law of the ultrasonic signal inside the transformer is obtained. Fabry–Pérot (F–P) fi bre acoustic sensors with a centre frequency of 28 kHz were fabricated. A partial discharge detection test was carried out in a 35 kV transformer winding model using the F–P sensors. The test results show that the ultrasonic waveform detected by the F–P sensors are in good agreement with the simulation results, and the propagation of the ultrasonic wave inside the transformer is verifi ed. It lays a foundation for detecting and locating PDs in power transformer by F–P acoustic sensors.

Relationship between perceived social support,self-efficiency and postpartum depression

GAO Yan,ZENG Juan,ZHANG Mei,ZHAO Zan-li,YANG Hua-li,HOU Li 한국간호과학회 2019 한국간호과학회 학술대회 Vol.2019 No.10

Research on Partial Discharge Pattern Recognition in GIS Based on EFPI Sensor

Wang Zan,Liu Zhongquan,Qiao Lili,Qian Dingdong,Chen Zhongxian,Gao Chaofei,Wang Wei 대한전기학회 2024 Journal of Electrical Engineering & Technology Vol.19 No.1

An EFPI fber optic ultrasonic sensor can be used for the detection and pattern recognition of partial discharge ultrasonic signals in Gas Insulated Switchgear. Compared with traditional piezoelectric sensors, it has many advantages, such as high sensitivity and strong anti-interference ability. Based on this, four typical PD models of the tip, metal particles, suspension and creeping surface were set up in the GIS cavity flled with 0.4 MPa SF6 gas, and the EFPI sensor was innovatively used to detect the discharge ultrasonic signal and extract the single ultrasonic pulse signal. The waveform features form a feature parameter database, and the probabilistic neural network algorithm and the support vector machine algorithm are used for pattern recognition and comparative analysis, respectively. The ultrasonic signal detected by the EFPI sensor has prominent features. On the basis of extracting the feature parameters, the two pattern recognition algorithms can achieve an average recognition rate of more than 85%, and the recognition efect of the support vector machine is better than that of the probabilistic neural network

Effects of chromium picolinate on fat deposition, activity and genetic expression of lipid metabolism-related enzymes in 21 day old Ross broilers

Chen, Guangxin,Gao, Zhenhua,Chu, Wenhui,Cao, Zan,Li, Chunyi,Zhao, Haiping Asian Australasian Association of Animal Productio 2018 Animal Bioscience Vol.31 No.4

Objective: This experiment was conducted to investigate the effects of chromium picolinate (CrP) on fat deposition, genetic expression and enzymatic activity of lipid metabolism-related enzymes. Methods: Two hundred forty one-day-old Ross broilers were randomly divided into 5 groups with 4 replicates per group and 12 Ross broiler chicks per replicate. The normal control group was fed a basal diet, and the other groups fed the same basal diet supplemented with 0.1, 0.2, 0.4, and 0.8 mg/kg CrP respectively. The experiment lasted for 21 days. Results: Added CrP in the basal diet decreased the abdominal fat, had no effects on subcutaneous fat thickness and inter-muscular fat width; 0.2 mg/kg CrP significantly decreased the fatty acid synthase (FAS) enzymatic (p<0.05); acetyl-CoA carboxylase (ACC) enzymatic activity decreased in all CrP groups (p<0.05); hormone-sensitive lipase (HSL) enzymatic activity also decreased, but the change was not significant (p>0.05); 0.4 mg/kg CrP group significantly decreased the lipoprotein lipase (LPL) enzymatic activity. FAS mRNA expression increased in all experimental groups, and the LPL mRNA expression significantly increased in all experimental groups (p<0.05), but not 0.2 mg/kg CrP group. Conclusion: The results indicated that adding CrP in basal diet decreased the abdominal fat percentage, had no effects on subcutaneous fat thickness and inter-muscular fat width, decreased the enzymatic activity of FAS, ACC, LPL and HSL and increased the genetic expression levels of FAS and LPL.

Three homologous genes encoding functional D8-sphingolipid desaturase in Populus tomentosa

Shu-Fen Li,Zan-Min Hu,Guo-Jun Zhang,Ying-Chun Yuan,Cong-Hui Wang,Wu-Jun Gao,Chuan-Liang Deng,Long-Dou Lu 한국유전학회 2014 Genes & Genomics Vol.36 No.3

Δ8-sphingolipid desaturase is characterized by itsability to catalyze desaturation at the C8 position of the longchainbase of sphingolipids in plants. No previous studieshave been conducted on genes encoding Δ8-sphingolipiddesaturases in the woody plant Populus tomentosa. In thisstudy, three genes that encode Δ8-sphingolipid desaturasewere isolated fromP. tomentosa. Among these genes, PtD8Aand PtD8B showed high sequence similarity; whereas PtD8Cexhibited large sequence divergence.RT-PCRresults showedthat PtD8A and PtD8B were expressed in all tissues detected,whereas PtD8C was not expressed in roots. Heterologousexpression in yeast revealed that PtD8A/B/C were functionalΔ8-sphingolipid desaturases, and can catalyze the C18-phytosphingeninedesaturation to produce 8(Z)- and 8(E)-C18-phytosphingenine.However, the conversion rate and ratios ofthe two products differed. Compared with control cells,transgenic yeasts expressing PtD8A/B/C exhibited enhancedaluminum tolerance. Our findings further elucidated thebiochemical functions and evolutionary history ofΔ8-sphingolipid desaturases in plants. Candidate genes forbreeding new poplar germplasm resources with enhancedtolerance ability to aluminium were also provided.

Immunomodulatory Properties of Lactobacillus plantarum NC8 Expressing an Anti-CD11c Single-Chain Fv Fragment

( Jing Liu ),( Guilian Yang ),( Xing Gao ),( Zan Zhang ),( Yang Liu ),( Xin Yang ),( Chunwei Shi ),( Qiong Liu ),( Yanlong Jiang ),( Chunfeng Wang ) 한국미생물 · 생명공학회 2019 Journal of microbiology and biotechnology Vol.29 No.1

The lactic acid bacteria species Lactobacillus plantarum (L. plantarum) has been used extensively for vaccine delivery. Considering to the critical role of dendritic cells in stimulating host immune response, in this study, we constructed a novel CD11c-targeting L. plantarum strain with surface-displayed variable fragments of anti-CD11c, single-chain antibody (scFv-CD11c). The newly designed L. plantarum strain, named 409-aCD11c, could adhere and invade more efficiently to bone marrow-derived DCs (BMDCs) in vitro due to the specific interaction between scFv-CD11c and CD11c located on the surface of BMDCs. After incubation with BMDCs, the 409-aCD11c strain harboring a eukaryotic vector pValac-GFP could lead to more efficient expression of GFP compared with wild-type strains shown by flow cytometry analysis, indicating the enhanced translocation of pValac-GFP from L. plantarum to BMDCs. Similar results were also observed in an in vivo study, which showed that oral administration resulted in efficient expression of GFP in both Peyer’s patches (PP) and mesenteric lymph nodes (MLNs) within 7 days after the last administration. In addition, the CD11c-targeting strain significantly promoted the differentiation and maturation of DCs, the differentiation of IL-4+ and IL-17A+ T helper (Th) cells in MLNs, as well as production of B220+ IgA+ B cells in the PP. In conclusion, this study developed a novel DC-targeting L. plantarum strain which could increase the ability to deliver eukaryotic expression plasmid to host cells, indicating a promising approach for vaccine study.

Outcomes of Microendoscopic Discectomy and Percutaneous Transforaminal Endoscopic Discectomy for the Treatment of Lumbar Disc Herniation: A Comparative Retrospective Study

Arjun Sinkemani,Xin Hong,Zeng-Xin Gao,Su-Yang Zhuang,Zan-Li Jiang,Shao-Dong Zhang,Jun-Ping Bao,Lei Zhu,Pei Zhang,Xin-Hui Xie,Feng Wang,Xiao-Tao Wu 대한척추외과학회 2015 Asian Spine Journal Vol.9 No.6

Study Design: Retrospective, case control evaluation of 86 patients who underwent microendoscopic discectomy (MED) and percutaneous transforaminal endoscopic discectomy (PTED) for the treatment of lumbar disc herniation (LDH). Purpose: To evaluate the safety and the outcomes of MED and PTED for the treatment of LDH. Overview of Literature: MED and PTED are minimally invasive surgical techniques for lower back pain. Studies to date have shown that MED and PTED are safe and effective treatment modalities for LDH. Methods: A retrospective study was performed in patients with LDH treated with MED (n=50) and transforaminal endoscopic discectomy (PTED; n=36) in our hospital. All patients were followed-up with self-evaluation questionnaires, Oswestry disability index (ODI), medical outcomes study 36-item short form health survey and MacNab criteria. All the patients in both groups were followed up to 12 months after the operation. Results: ODI questionnaire responses were not statistically different between the MED and PTED groups (53.00 vs. 48.72) before treatment. Average scores and minimal disability after 5 days to 12 months of follow-up were 4.96 in the MED group and 3.61 in the PTED group. According to MacNab criteria, 92.0% of the MED group and 94.4% of the PTED group had excellent or good results with no significant difference. Conclusions: There was no significant difference between MED and PTED outcomes. Further large-scale, randomized studies with long-term follow-up are needed.