Human skeleton contains significant information about actions, therefore, it is quite intuitive to incorporate skeletons in human action recognition. Human skeleton resembles to a graph where body ...joints and bones mimic to graph nodes and edges. This resemblance of human skeleton to graph structure is the main motivation to apply graph convolutional neural network for human action recognition. Results show that the discriminant contribution of different joints is not equal for different actions. Therefore, we propose to use attention-joints that correspond to joints significantly contributing to the specific actions. Features corresponding to only these attention-joints are computed and assigned as node features of the graph. In our method, node features (also termed as attention-joint features) include the i) distances of attention-joints from the center-of-gravity of human body, ii) distances between adjacent attention-joints and iii) joints flow features. The proposed method gives a simple but more efficient representation of skeleton sequences by concatenating more relative distances and relative coordinates to other joints. The proposed methodology has been evaluated on single image Stanford 40-Actions dataset, as well as on temporal skeleton-based action recognition PKU-MDD and NTU-RGBD datasets. Results show that this framework outperforms existing state-of-the-art methods.
Heteroaryldihydropyrimidine (HAP) and sulfamoylbenzamide (SBA) are promising non-nucleos(t)ide HBV replication inhibitors. HAPs are known to promote core protein mis-assembly, but the molecular ...mechanism of abnormal assembly is still elusive. Likewise, the assembly status of core protein induced by SBA remains unknown. Here we show that SBA, unlike HAP, does not promote core protein mis-assembly. Interestingly, two reference compounds HAP_R01 and SBA_R01 bind to the same pocket at the dimer-dimer interface in the crystal structures of core protein Y132A hexamer. The striking difference lies in a unique hydrophobic subpocket that is occupied by the thiazole group of HAP_R01, but is unperturbed by SBA_R01. Photoaffinity labeling confirms the HAP_R01 binding pose at the dimer-dimer interface on capsid and suggests a new mechanism of HAP-induced mis-assembly. Based on the common features in crystal structures we predict that T33 mutations generate similar susceptibility changes to both compounds. In contrast, mutations at positions in close contact with HAP-specific groups (P25A, P25S, or V124F) only reduce susceptibility to HAP_R01, but not to SBA_R01. Thus, HAP and SBA are likely to have distinctive resistance profiles. Notably, P25S and V124F substitutions exist in low-abundance quasispecies in treatment-naïve patients, suggesting potential clinical relevance.
Video-based human action recognition is one of the most important and challenging areas of research in the field of computer vision. Human action recognition has found many pragmatic applications in ...video surveillance, human-computer interaction, entertainment, autonomous driving, etc. Owing to the recent development of deep learning methods for human action recognition, the performance of action recognition has significantly enhanced for challenging datasets. Deep learning techniques are mainly used for recognizing actions in images and videos comprising of Euclidean data. A recent development in deep learning methods is the extension of these techniques to non-Euclidean data or graph data with many nodes and edges. Human body skeleton resembles a graph, therefore, the graph convolutional network (GCN) is applicable to the non-Euclidean body skeleton. In the past few years, GCN has emerged as an important tool for skeleton-based action recognition. Therefore, we conduct a survey using GCN methods for action recognition. Herein, we present a comprehensive overview of recent GCN techniques for action recognition, propose a taxonomy for the categorization of GCN techniques for action recognition, carry out a detailed study of the benchmark datasets, enlist relevant resources and open-source codes, and finally provide an outline for future research directions and trends. To the best of authors' knowledge, this is the first survey for action recognition using GCN techniques.
Impact Statement- Graph convolutional neural networks have marked a great progress in recent years. There is a similarity between body skeleton and a graph; therefore, GCNs have been widely used for skeleton-based action recognition. In this article, we summarize recent graph-based action recognition techniques, provide a deeper insight of these methods, list source-codes and available resources. This article will help the researchers to develop a basic understanding of graph convolutional methods for action recognition, benefit from useful resources, and think about future directions.
Recently, deep learning has greatly promoted the performance of license plate recognition (LPR) by learning robust features from numerous labeled data. However, the large variation of wild license ...plates across complicated environments and perspectives is still a huge challenge to the robust LPR. To solve the problem, we propose an effective and efficient shared adversarial training network (SATN) in this paper, which can learn the environment-independent and perspective-free semantic features from wild license plates with the prior knowledge of standard stencil-rendered license plates, as standard stencil-rendered license plates are independent of complicated environments and various perspectives. Besides, to correct the features of heavily perspective distorted license plates perfectly, we further propose a novel dual attention transformation (DAT) module in the shared adversarial training network. Comprehensive experiments on AOLP-RP and CCPD benchmarks show that the proposed method outperforms state-of-the-art methods by a large margin on the LPR task.
Inhibition of embryonic ectoderm development (EED) is a new cancer therapeutic strategy. Herein, we report our discovery of EEDi-5285 as an exceptionally potent, efficacious, and orally active EED ...inhibitor. EEDi-5285 binds to the EED protein with an IC50 value of 0.2 nM and inhibits cell growth with IC50 values of 20 pM and 0.5 nM in the Pfeiffer and KARPAS422 lymphoma cell lines, respectively, carrying an EZH2 mutation. EEDi-5285 is approximately 100 times more potent than EED226 in binding to EED and >300 times more potent than EED226 in inhibition of cell growth in the KARPAS422 cell line. EEDi-5285 has excellent pharmacokinetics and achieves complete and durable tumor regression in the KARPAS422 xenograft model in mice with oral administration. The cocrystal structure of EEDi-5285 in a complex with EED defines the precise structural basis for their high binding affinity. EEDi-5285 is the most potent and efficacious EED inhibitor reported to date.
Embryonic ectoderm development (EED) is a promising therapeutic target for human cancers and other diseases. We report herein the discovery of exceptionally potent and efficacious EED inhibitors. By ...conformational restriction of a previously reported EED inhibitor, we obtained a potent lead compound. Further optimization of the lead yielded exceptionally potent EED inhibitors. The best compound EEDi-5273 binds to EED with an IC50 value of 0.2 nM and inhibits the KARPAS422 cell growth with an IC50 value of 1.2 nM. It demonstrates an excellent PK and ADME profile, and its oral administration leads to complete and persistent tumor regression in the KARPAS422 xenograft model with no signs of toxicity. Co-crystal structures of two potent EED inhibitors with EED provide a solid structural basis for their high-affinity binding. EEDi-5273 is a promising EED inhibitor for further advanced preclinical development for the treatment of human cancer and other human diseases.
Targeting the capsid protein of hepatitis B virus (HBV) and thus interrupting normal capsid formation have been an attractive approach to block the replication of HBV viruses. We carried out ...multidimensional structural optimizations based on the heteroaryldihydropyrimidine (HAP) analogue Bay41-4109 (1) and identified a novel series of HBV capsid inhibitors that demonstrated promising cellular selectivity indexes, metabolic stabilities, and in vitro safety profiles. Herein we disclose the design, synthesis, structure–activity relationship (SAR), cocrystal structure in complex with HBV capsid proteins and in vivo pharmacological study of the 4-methyl HAP analogues. In particular, the (2S,4S)-4,4-difluoroproline substituted analogue 34a demonstrated high oral bioavailability and liver exposure and achieved over 2 log viral load reduction in a hydrodynamic injected (HDI) HBV mouse model.
Action recognition has been achieved great progress in recent years because of better feature representation learning and classification technology like convolutional neural networks (CNNs). However, ...most current deep learning approaches treat the action recognition as a black box, ignoring the specific domain knowledge of action itself. In this paper, by analyzing the characteristics of different actions, we proposed a new framework that involves residual-attention module and joint path-signature feature (JPSF) representation framework. The path signature theory was developed recently in the field of rough path and stochastic analysis, which provides a very efficient way to analyze any temporal sequence data. The proposed n-fold joint path signature features entail the Euclidean distances between joints and respective angles. For our experiment, JPSF for three modalities of joints (spatial location, bi-folds and tri-folds) are computed over the temporal length of action sequence. Then all these PSF are concatenated and fed to a CNN to give the recognition result. Experiments on three benchmark datasets, J-HMDB, HMDB-51 and UCF-101, indicate that our proposed method achieves state-of-the-art performance.
A structure-based approach was employed to design a new class of small-molecule inhibitors of Bcl-2. The most potent compound 5 (TW-37) binds to Bcl-2 with a K i value of 290 nM and also to Bcl-xL ...and Mcl-1 with high affinities. Compound 5 potently inhibits cell growth in PC-3 prostate cancer cells with an IC50 value of 200 nM and effectively induces apoptosis in a dose-dependent manner.
A series of acylpyrogallols were designed, synthesized, and evaluated as small-molecule inhibitors of antiapoptotic Bcl-2 proteins. The most potent compound 9 (TM-179) binds to Bcl-2 with an IC50 of ...170 nM and to Mcl-1 with a K i of 37 nM. Compound 9 potently inhibits cell growth and induces apoptosis in human breast and prostate cancer cell lines.