Explore the vast range of titles available.

Multi-object tracking (MOT) is an important problem in computer vision which has a wide range of applications. Formulating MOT as multi-task learning of object detection and re-ID in a single network is appealing since it allows joint optimization of the two tasks and enjoys high computation efficiency. However, we find that the two tasks tend to compete with each other which need to be carefully addressed. In particular, previous works usually treat re-ID as a secondary task whose accuracy is heavily affected by the primary detection task. As a result, the network is biased to the primary detection task which is not fair to the re-ID task. To solve the problem, we present a simple yet effective approach termed as FairMOT based on the anchor-free object detection architecture CenterNet. Note that it is not a naive combination of CenterNet and re-ID. Instead, we present a bunch of detailed designs which are critical to achieve good tracking results by thorough empirical studies. The resulting approach achieves high accuracy for both detection and tracking. The approach outperforms the state-of-the-art methods by a large margin on several public datasets. The source code and pre-trained models are released at https://github.com/ifzhang/FairMOT.

Occlusion is probably the biggest challenge for human pose estimation in the wild. Typical solutions often rely on intrusive sensors such as IMUs to detect occluded joints. To make the task truly unconstrained, we present AdaFuse, an adaptive multiview fusion method, which can enhance the features in occluded views by leveraging those in visible views. The core of AdaFuse is to determine the point-point correspondence between two views which we solve effectively by exploring the sparsity of the heatmap representation. We also learn an adaptive fusion weight for each camera view to reflect its feature quality in order to reduce the chance that good features are undesirably corrupted by “bad” views. The fusion model is trained end-to-end with the pose estimation network, and can be directly applied to new camera configurations without additional adaptation. We extensively evaluate the approach on three public datasets including Human3.6M, Total Capture and CMU Panoptic. It outperforms the state-of-the-arts on all of them. We also create a large scale synthetic dataset Occlusion-Person, which allows us to perform numerical evaluation on the occluded joints, as it provides occlusion labels for every joint in the images. The dataset and code are released at https://github.com/zhezh/adafuse-3d-human-pose.

The molecular weight, the triple-helix conformation, the monosaccharide content, the manner of glycosidic linkages, and the polysaccharide conjugates of polysaccharides all affect bioactivity. The purpose of this study was to determine how different molecular weights affected the bioactivity of the Lycium barbarum polysaccharides (LBPs). By ethanol-graded precipitation and ultrafiltration membrane separation, one oligosaccharide (LBPs-1, 1.912 kDa) and two polysaccharides (LBPs-2, 7.481 kDa; LBPs-3, 46.239 kDa) were obtained from Lycium barbarum. While the major component of LBPs-1 and LBPs-2 was glucose, the main constituents of LBPs-3 were arabinose, galactose, and glucose. LBPs-2 and LBPs-3 exhibited triple-helix conformations, as evidenced by the Congo red experiment and AFM data. Sugar residues of LBPs-2 and LBPs-3 were elucidated by NMR spectra. The polysaccharides (LBPs-2 and LBPs-3) exhibited much higher antioxidant capacities than oligosaccharide (LBPs-1). LBPs-3 showed higher oxygen radical absorbance capacity (ORAC) and superoxide dismutase (SOD) activity than LBPs-2, but a lower capability for scavenging ABTS+ radicals. In zebrafish, LBPs-2 and LBPs-3 boosted the growth of T-lymphocytes and macrophages, enhanced the immunological response, and mitigated the immune damage generated by VTI. In addition to the molecular weight, the results indicated that the biological activities would be the consequence of various aspects, such as the monosaccharide composition ratio, the chemical composition, and the chemical reaction mechanism.

Store-operated calcium channels (SOCCs) are involved in the process of cell proliferation; however, their expression levels differ among cell types and information on their effects in different cells is lacking. This study aimed to compare the differing effects of SOCCs on the proliferation of vascular smooth muscle cells (VSMCs) and vascular endothelial cells (VECs), and the repair ability of SOCC after vascular endothelial injury. Rat primary coronary VSMCs and VECs were cultured in vitro and expression levels of SOCC molecules were detected by western blotting and quantitative polymerase chain reaction. Various molecules were selected and transfected into VSMCs and VECs using an adenovirus vector, and cell proliferation, the cell cycle, and intracellular Ca2+ were then detected. We also established a rat carotid artery endothelial injury model to verify the results of the in vitro experiments. Expression levels of transient receptor potential canonical (TRPC) 3 and TRPC5 were higher in VSMCs than in VECs. Silencing TRPC3/5 significantly inhibited cell proliferation and Ca2+ influx in VSMCs, but not in VECs. Silencing TRPC3/5 after rat carotid artery endothelial injury inhibited neointima formation, with a better reparative effect on the endothelial cell layer than rapamycin. TRPC3/5 participates in the proliferation of VSMCs via SOCCs, and silencing its expression inhibits the formation of neointima after endothelial injury. However, this effect was not significant in VECs, suggesting that other compensatory pathways may have emerged.

Security is a major concern in an increasingly multimedia-defined universe where the Internet serves as an indispensable resource for information and entertainment. Digital Rights Management (DRM) is the technology by which network systems protect and provide access to critical and time-sensitive copyrighted material and/or personal information. This book equips savvy technology professionals and their aspiring collegiate protégés with the latest technologies, strategies and methodologies needed to successfully thwart off those who thrive on security holes and weaknesses. Filled with sample application scenarios and algorithms, this book provides an in-depth examination of present and future field technologies including encryption, authentication, copy control, tagging, tracing, conditional access and media identification. The authors present a diversified blend of theory and practice and focus on the constantly changing developments in multimedia applications thus providing an admirably comprehensive book. * Discusses state-of-the-art multimedia authentication and fingerprinting techniques * Presents several practical methodologies from industry, including broadcast encryption, digital media forensics and 3D mesh watermarking * Focuses on the need for security in multimedia applications found on computer networks, cell phones and emerging mobile computing devices

Image hash functions find extensive applications in content authentication, database search. This paper develops a novel algorithm for generating a secure and robust image hash based on compressive sensing and Fourier-Mellin transform. Firstly, we incorporate Fourier-Mellin transform into our method to improve its performance under rotation, scale, transition attacks. Secondly, we exploit the property of dimension reduction inherent in compressive sensing for hash design. The statistic structure and sparse of the wavelet coefficients assure efficient compression in situation of including maximum the image features. The hashing method is computationally secure without additional randomization process. Such a combined approach is capable of tackling all types of attacks and thus can yield a better overall performance in multimedia identification. To demonstrate the superior performance of the proposed schemes, receiver operating characteristics analysis over a large image database is performed. Experimental results show that the proposed image hashing is robust to a wide range of distortions and attacks. When compared with the current state-of-the-art methods, the proposed method yields better identification performances under geometric attacks such as rotation attacks and brightness changes.

Three-color electrophoretic displays (EPDs) are a new type of optoelectronic display device. However, they have the defect of red ghost images during gray scale transformation, which affects the accuracy of the gray scale display. In this paper, we proposed a new driving method for eliminating the red ghost images. A driving waveform was composed of an erasing stage, an activation stage, and a driving stage. First, the erasing stage was subdivided into a red erasing stage and an original erasing stage, the red erasing stage was used to eliminate residual red particles in the top of the microcapsules. Then, a high-frequency square wave was used as the activation stage for increasing the activity of the black and white particles. Meanwhile, the intensity of flickers could be decreased by the high-frequency square wave. Finally, the performance of the driving waveform was tested by a colorimeter. The experimental results showed that the driving waveform could effectively eliminate red ghost images by 80.43% and reduce the flicker intensity by 79.63%, compared with an existing driving waveform.

The cytoskeleton is composed of microtubules, microfilaments, and intermediate filaments in cells. While the functions of microtubules and microfilaments have been well elucidated, the roles of intermediate filaments and associated proteins remain largely unknown, especially in meiosis. BFSP1 is an intermediate filament protein mainly expressed in the eye lens to play important roles in the development of congenital cataract. Here, we document that BFSP1 functions as a spindle regulator to drive the oocyte asymmetric division. Specifically, we found that BFSP1 distributed on the spindle apparatus during oocyte meiotic maturation. Depletion of BFSP1 resulted in symmetric division of oocytes, accompanied by the formation of elongated spindles at metaphase I and anaphase/telophase I stages. In addition, immunoprecipitation combined with mass spectrometry analysis identified MAP1B, a microtubule‐associated protein, as an interacting partner of BFSP1. Depletion or mutation of MAP1B phenocopied the meiotic defects observed in BFSP1‐depleted oocytes, and expression of exogenous MAP1B‐EGFP in BFSP1‐depleted oocytes recovered the spindle length and asymmetric division. We further determined that BFSP1 recruited molecular chaperone HSP90α on the spindle to stabilize MAP1B, thereby controlling the spindle length. To sum up, our findings reveal a unique meiotic role for BFSP1 in the regulation of spindle dynamics and oocyte asymmetric division. Different from mitosis, the female meiosis undergoes asymmetric division that produces haploid oocytes and polar body, which is essential for retaining maternal components to support subsequent fertilization and embryo development. However, the underlying mechanisms are still largely unknown. This work documents that an intermediate filament orphan protein BFSP1, predominantly expressed in lens cells, drives oocyte asymmetric division by modulating spindle length.

At present, three-color electrophoretic displays (EPDs) have problems of dim brightness and insufficient color saturation. In this paper, a driving waveform based on a damping oscillation was proposed to optimize the red saturation in three-color EPDs. The optimized driving waveform was composed of an erasing stage, a particles activation stage, a red electrophoretic particles purification stage, and a red display stage. The driving duration was set to 360 ms, 880 ms, 400 ms, and 2400 ms, respectively. The erasing stage was used to erase the current pixel state and refresh to a black state. The particles’ activation stage was set as two cycles, and then refreshed to the black state. The red electrophoretic particles’ purification stage was a damping oscillation driving waveform. The red and black electrophoretic particles were separated by changing the magnitude and polarity of applied electric filed, so that the red electrophoretic particles were purified. The red display stage was a low positive voltage, and red electrophoretic particles were driven to the common electrode to display a red state. The experimental results showed that the maximum red saturation could reach 0.583, which was increased by 27.57% compared with the traditional driving waveform.

We recently demonstrated that retinoic acid receptor-γ (RARγ) is overexpressed and acts as a tumor promoter in hepatocellular carcinoma (HCC). The oncogenic activity of RARγ is mainly attributed to its physiological interaction with p85α regulatory subunit of PI3K leading to constitutive activation of AKT. Here we report RARγ as a negative regulator of p53 signaling and thus extend the oncogenic potential of RARγ to a new role in controlling the balance between AKT and p53. A natural flavonoid acacetin is then identified to be capable of modulating RARγ-dependent AKT-p53 network. It specifically binds to RARγ and inhibits all-trans retinoic acid ( at RA) stimulation of RARγ transactivation. However, the anticancer action of acacetin is independent on its modulation of RARγ-driven transcriptional activity. Acacetin induces cancer cell apoptosis through antagonizing the non-genomic effect of RARγ on AKT and p53. When bound to RARγ, acacetin prevents RARγ from its activation of AKT followed by recovery of the normal p53 signaling. Given the implication of AKT-p53 dysregulation in most HCC, targeting the non-genomic signaling of RARγ that switches AKT-p53 from a pro-survival to a pro-apoptotic program in cancer cells should be a promising strategy for developing novel anti-HCC drugs.