Open-Ended Online Learning for Autonomous Visual Perception.

The visual perception systems aim to autonomously collect consecutive visual data and perceive the relevant information online like human beings. In comparison with the classical static visual systems focusing on fixed tasks (e.g., face recognition for visual surveillance), the real-world visual systems (e.g., the robot visual system) often need to handle unpredicted tasks and dynamically changed environments, which need to imitate human-like intelligence with open-ended online learning ability. Therefore, we provide a comprehensive analysis of open-ended online learning problems for autonomous visual perception in this survey. Based on "what to online learn" among visual perception scenarios, we classify the open-ended online learning methods into five categories: instance incremental learning to handle data attributes changing, feature evolution learning for incremental and decremental features with the feature dimension changed dynamically, class incremental learning and task incremental learning aiming at online adding new coming classes/tasks, and parallel and distributed learning for large-scale data to reveal the computational and storage advantages. We discuss the characteristic of each method and introduce several representative works as well. Finally, we introduce some representative visual perception applications to show the enhanced performance when using various open-ended online learning models, followed by a discussion of several future directions.

Oil body extraction from high-fat and high-protein soybeans by laccase cross-linked beet pectin: physicochemical and oxidation properties.

BACKGROUND: Soybean oil bodies (SOB) are droplets of natural emulsified oil. Soybean oil emulsifies well but it is easily oxidized during storage. Beet pectin is a complex anionic polysaccharide, which can be adsorbed on the surface of liposomes to improve their resistance to flocculation. Laccase can covalently cross-link ferulic acid in beet pectin, and its structure is irreversible, which can improve the stability of polysaccharides. RESULTS: At pH 2.5, laccase cross-linked beet pectin high-oil soybean oil body (HOSOB) and high-protein soybean oil body (HPSOB) emulsions showed obvious aggregation and severe stratification, and the oxidation of the emulsions was also high. The flocculation of emulsions decreased with an increase in the pH. The effect of pH on the flocculation of emulsion was confirmed by confocal laser electron microscopy. The ζ potential, emulsification, and rheological shear force increased with increasing pH whereas the particle size and surface hydrophobicity decreased with increasing pH. CONCLUSION: This experiment indicates that the physicochemical stability of the two composite emulsions was strongly affected under acidic conditions but stable under neutral and weakly alkaline conditions. Under the same acid-base conditions, the degree of oxidation of HPSOB composite emulsion changes substantially. The results of this study can provide a basis for the design of very stable emulsions to meet the demand for natural products. © 2023 Society of Chemical Industry.

A Comprehensive Study of 3-D Vision-Based Robot Manipulation.

Robot manipulation, for example, pick-and-place manipulation, is broadly used for intelligent manufacturing with industrial robots, ocean engineering with underwater robots, service robots, or even healthcare with medical robots. Most traditional robot manipulations adopt 2-D vision systems with plane hypotheses and can only generate 3-DOF (degrees of freedom) pose accordingly. To mimic human intelligence and endow the robot with more flexible working capabilities, 3-D vision-based robot manipulation has been studied. However, this task is still challenging in the open world especially for general object recognition and pose estimation with occlusion in cluttered backgrounds and human-like flexible manipulation. In this article, we propose a comprehensive analysis of recent progress about the 3-D vision for robot manipulation, including 3-D data acquisition and representation, robot-vision calibration, 3-D object detection/recognition, 6-DOF pose estimation, grasping estimation, and motion planning. We then present some public datasets, evaluation criteria, comparisons, and challenges. Finally, the related application domains of robot manipulation are given, and some future directions and open problems are studied as well.

Risk factors for respiratory failure after tuberculosis-destroyed lung surgery and increased dyspnea score at 1-year follow-up.

Background: Although pneumonectomy is an important surgical treatment for tuberculosis-destroyed lung (TDL), few studies have investigated long-term postoperative TDL prognosis. Here, risk factors were determined for postoperative secondary respiratory failure and modified British Medical Research Council (mMRC ≥1) at discharge and at 1-year post-surgical follow-up. Methods: A two-way cohort study was conducted of 116 patients admitted to our thoracic surgery department for surgical TDL treatment from January 2001 to June 2020. General clinical data were collected then patient postoperative mMRC scores were monitored for 1 year. Dyspnea-associated factors (mMRC ≥1) were identified then risk factors for postoperative respiratory failure and compromised long-term respiratory function were identified using multivariate adjusted logistic regression analysis. Results: Of 116 patients, 27.6% (32/116) developed respiratory failure secondary to surgery. Multifactorial logistic regression analysis revealed that preoperative serum albumin of <30 g/L [adjusted odds ratios (aOR) 6.613, 95% confidence intervals (CI): 1.064-41.086] and intraoperative bleeding of >1,000 mL (aOR 6.876, 95% CI: 1.236-38.243) were risk factors for subsequent respiratory failure only in patients experiencing postoperative secondary respiratory failure. Sorting of patient mMRC dyspnea index scores into two groups (mMRC =0, mMRC ≥1) followed by logistic regression analysis revealed that risk factors for 1-year postoperative dyspnea included mMRC score ≥1 at discharge (aOR 14.446, 95% CI: 1.102-189.361) and postoperative respiratory failure occurrence (aOR 9.946, 95% CI: 1.063-93.034). Conclusions: TDL patient preoperative hypoalbuminemia and extensive intraoperative bleeding were risk factors for postoperative secondary respiratory failure. Postoperative secondary respiratory failure and high mMRC (≥1) at discharge were associated with reduced postoperative long-term recovery of respiratory function.

Long-term follow-up of tuberculosis-destroyed lung patients after surgical treatment.

BACKGROUND: To monitor dypsnea and mortality at 5 and 10 years, respectively, after surgical treatment of tuberculosis-destroyed lung (TDL) patients. METHODS: TDL patients treated surgically at Beijing Chest Hospital from November 2007 to June 2019 were monitored in this observational study. Follow-up assessments of respiratory function indicators and survival conducted 5 and 10 years post-surgery led to patient grouping based on mMRC score into a dyspnea group (mMRC ≥ 1) and a non-dyspnea group (mMRC = 0). Cox regression analysis detected effects of patient demographics, clinical characteristics, surgical factors and respiratory function on 5 year post-surgical survival. RESULTS: By study completion (June 30, 2020), 32 of 104 patients were lost and 72 completed follow-up for a study total of 258.9 person-years. 45 patients (62.5%, 45/72) had mMRC scores of 0, while 12 (16.7%, 12/72), 21 (36.2%, 21/58) and 27 (60.0%, 27/45) patients exhibited dyspnea by 1, 3 and 5 years post-surgery, respectively. Low lung carbon monoxide diffusion score (DLCO% pred) and scoliosis contributed to dyspnea occurrence. CONCLUSIONS: Most TDL patients lacked subjective dyspnea signs post-surgery, while dyspnea rates increased with time. Preoperative low lung diffusion function and Scoliosis were associated with factors for postoperative dyspnea. Surgical treatment increased TDL patient survival overall.

An Internet-Oriented Multilayer Network Model Characterization and Robustness Analysis Method.

The Internet creates multidimensional and complex relationships in terms of the composition, application and mapping of social users. Most of the previous related research has focused on the single-layer topology of physical device networks but ignored the study of service access relationships and the social structure of users on the Internet. Here, we propose a composite framework to understand how the interaction between the physical devices network, business application network, and user role network affects the robustness of the entire Internet. In this paper, a multilayer network consisting of a physical device layer, business application layer and user role layer is constructed by collecting experimental network data. We characterize the disturbance process of the entire multilayer network when a physical entity device fails by designing nodal disturbance to investigate the interactions that exist between the different network layers. Meanwhile, we analyze the characteristics of the Internet-oriented multilayer network structure and propose a heuristic multilayer network topology generation algorithm based on the initial routing topology and networking pattern, which simulates the evolution process of multilayer network topology. To further analyze the robustness of this multilayer network model, we combined a total of six target node ranking indicators including random strategy, degree centrality, betweenness centrality, closeness centrality, clustering coefficient and network constraint coefficient, performed node deletion simulations in the experimental network, and analyzed the impact of component types and interactions on the robustness of the overall multilayer network based on the maximum component change in the network. These results provide new insights into the operational processes of the Internet from a multi-domain data fusion perspective, reflecting that the coupling relationships that exist between the different interaction layers are closely linked to the robustness of multilayer networks.

Oil bodies extracted from high-oil soybeans (Glycine max) exhibited higher oxidative and physical stability than oil bodies from high-protein soybeans.

Reports concerning the characteristics of soybean oil bodies (SOBs) isolated from high protein genotypes and high oil genotypes of soybeans available in the literature are insufficient and limiting. In this study, fatty acid compositions, total phenol and tocopherol contents, antioxidant capacity, and physicochemical stability of SOB emulsions recovered from three high-protein and three high-oil genotype soybeans were comparatively investigated. Principal component analysis showed that all six SOB samples could be easily discriminated based on the cultivar characteristics. Overall, the SOBs derived from the high-protein soybeans exhibited higher polyunsaturated fatty acid (PUFA) contents, while the SOBs derived from the high-oil soybeans had higher extraction yields and tocopherol contents; the tocopherol content was also positively correlated with the antioxidant capacity of the lipophilic fraction, but the difference in the total phenolic content between the two genotypes was not significant. The SOBs derived from the high-protein soybeans were more easily oxidized during storage, with 1.38- and 4-fold higher accumulation rates of lipid hydroperoxides (LPO) and thiobarbituric acid reactive substances (TBARS), respectively, in the high-protein-derived SOBs than in the high-oil-derived SOBs. In addition, the SOBs from the high-protein soybeans exhibited pronounced coalescence during storage, which was corroborated by focused confocal microscopy. These results confirmed that SOBs obtained from high-oil soybean genotypes are more suitable to manufacture OB-based products due to their superior physicochemical stability.

Reversible Data Hiding under Inconsistent Distortion Metrics.

Recursive code construction (RCC), based on the optimal transition probability matrix (OTPM), approaching the rate-distortion bound of reversible data hiding (RDH) has been proposed. Using the existing methods, OTPM can be effectively estimated only for a consistent distortion metric, i.e., if the host elements at different positions share the same distortion metric. However, in many applications, the distortion metrics are position dependent and should thus be inconsistent. Inconsistent distortion metrics can usually be quantified as a multi-distortion metric. In this paper, we first formulate the rate-distortion problem of RDH under a multi-distortion metric and subsequently propose a general framework to estimate the corresponding OTPM, with which RCC is extended to approach the rate-distortion bound of RDH under the multi-distortion metric. We apply the proposed framework to two examples of inconsistent distortion metrics: RDH in color image and reversible steganography. The experimental results show that the proposed method can efficiently improve upon the existing techniques.