Tailored multi-organ segmentation with model adaptation and ensemble.

Multi-organ segmentation, which identifies and separates different organs in medical images, is a fundamental task in medical image analysis. Recently, the immense success of deep learning motivated its wide adoption in multi-organ segmentation tasks. However, due to expensive labor costs and expertise, the availability of multi-organ annotations is usually limited and hence poses a challenge in obtaining sufficient training data for deep learning-based methods. In this paper, we aim to address this issue by combining off-the-shelf single-organ segmentation models to develop a multi-organ segmentation model on the target dataset, which helps get rid of the dependence on annotated data for multi-organ segmentation. To this end, we propose a novel dual-stage method that consists of a Model Adaptation stage and a Model Ensemble stage. The first stage enhances the generalization of each off-the-shelf segmentation model on the target domain, while the second stage distills and integrates knowledge from multiple adapted single-organ segmentation models. Extensive experiments on four abdomen datasets demonstrate that our proposed method can effectively leverage off-the-shelf single-organ segmentation models to obtain a tailored model for multi-organ segmentation with high accuracy.

InOR-Net: Incremental 3-D Object Recognition Network for Point Cloud Representation.

3-D object recognition has successfully become an appealing research topic in the real world. However, most existing recognition models unreasonably assume that the categories of 3-D objects cannot change over time in the real world. This unrealistic assumption may result in significant performance degradation for them to learn new classes of 3-D objects consecutively due to the catastrophic forgetting on old learned classes. Moreover, they cannot explore which 3-D geometric characteristics are essential to alleviate the catastrophic forgetting on old classes of 3-D objects. To tackle the above challenges, we develop a novel Incremental 3-D Object Recognition Network (i.e., InOR-Net), which could recognize new classes of 3-D objects continuously by overcoming the catastrophic forgetting on old classes. Specifically, category-guided geometric reasoning is proposed to reason local geometric structures with distinctive 3-D characteristics of each class by leveraging intrinsic category information. We then propose a novel critic-induced geometric attention mechanism to distinguish which 3-D geometric characteristics within each class are beneficial to overcome the catastrophic forgetting on old classes of 3-D objects while preventing the negative influence of useless 3-D characteristics. In addition, a dual adaptive fairness compensations' strategy is designed to overcome the forgetting brought by class imbalance by compensating biased weights and predictions of the classifier. Comparison experiments verify the state-of-the-art performance of the proposed InOR-Net model on several public point cloud datasets.

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.

PFDN: Pyramid Feature Decoupling Network for Single Image Deraining.

Restoring images degraded by rain has attracted more academic attention since rain streaks could reduce the visibility of outdoor scenes. However, most existing deraining methods attempt to remove rain while recovering details in a unified framework, which is an ideal and contradictory target in the image deraining task. Moreover, the relative independence of rain streak features and background features is usually ignored in the feature domain. To tackle these challenges above, we propose an effective Pyramid Feature Decoupling Network (i.e., PFDN) for single image deraining, which could accomplish image deraining and details recovery with the corresponding features. Specifically, the input rainy image features are extracted via a recurrent pyramid module, where the features for the rainy image are divided into two parts, i.e., rain-relevant and rain-irrelevant features. Afterwards, we introduce a novel rain streak removal network for rain-relevant features and remove the rain streak from the rainy image by estimating the rain streak information. Benefiting from lateral outputs, we propose an attention module to enhance the rain-irrelevant features, which could generate spatially accurate and contextually reliable details for image recovery. For better disentanglement, we also enforce multiple causality losses at the pyramid features to encourage the decoupling of rain-relevant and rain-irrelevant features from the high to shallow layers. Extensive experiments demonstrate that our module can well model the rain-relevant information over the domain of the feature. Our framework empowered by PFDN modules significantly outperforms the state-of-the-art methods on single image deraining with multiple widely-used benchmarks, and also shows superiority in the fully-supervised domain.

High-throughput sequencing reveals omnivorous and preferential diets of the rotifer Polyarthra in situ.

Knowledge of in situ diet of widespread rotifers is crucial for accurately understanding the trophic position, ecological function, and adaptability to environmental changes in aquatic ecosystems. However, it is challenging to achieve the in situ diet information due to the lack of efficient and comprehensive methods. Here, we investigated the diet composition of Polyarthra in a subtropical lake using high-throughput sequencing (HTS) of a rRNA metabarcode for Polyarthra and ambient water samples. After eliminating Polyarthra sequences, a total of 159 operational taxonomic units (OTUs) from taxa in 15 phyla were detected from Polyarthra gut content samples. Most of the OTUs belong to Chlorophyta, followed by unclassified Fungi, Chrysophyta, Dinoflagellata, Ciliophora, Bacillariophyta, Cryptophyta, Arthropoda, Cercozoa, Mollusca, Apicomplexa, Haptophyta, Amoebozoa, Chordata and other eukaryotes. Our results showed that Polyarthra mainly grazed on Chlorophyta, which may result from the high relative abundance of Chlorophyta in ambient waters. In contrast, Chrysophyceae and Synurophyceae were enriched in Polyarthra's gut, indicating that this rotifer prefers these taxa as food. Moreover, correlation analysis showed that total nitrogen, transparency, depth, Chlorophyll-a and total phosphorus were key factors for the variation of the eukaryotic community in the Polyarthra gut contents. When the concentration of nutrients in the water environment decreased, Polyarthra shifted from herbivorous feeding to more carnivorous feeding. Thus, Polyarthra is generally omnivorous but preference for Chrysophytes and Synurophytes, and it responds to the environmental changes by adopting a flexible feeding strategy. This could partly explain why the widespread rotifers have apparently wide tolerance toward spatial and environmental changes.

Visual-Tactile Fused Graph Learning for Object Clustering.

Object clustering has received considerable research attention most recently. However, 1) most existing object clustering methods utilize visual information while ignoring important tactile modality, which would inevitably lead to model performance degradation and 2) simply concatenating visual and tactile information via multiview clustering method can make complementary information to not be fully explored, since there are many differences between vision and touch. To address these issues, we put forward a graph-based visual-tactile fused object clustering framework with two modules: 1) a modality-specific representation learning module MR and 2) a unified affinity graph learning module MU . Specifically, MR focuses on learning modality-specific representations for visual-tactile data, where deep non-negative matrix factorization (NMF) is adopted to extract the hidden information behind each modality. Meanwhile, we employ an autoencoder-like structure to enhance the robustness of the learned representations, and two graphs to improve its compactness. Furthermore, MU highlights how to mitigate the differences between vision and touch, and further maximize the mutual information, which adopts a minimizing disagreement scheme to guide the modality-specific representations toward a unified affinity graph. To achieve ideal clustering performance, a Laplacian rank constraint is imposed to regularize the learned graph with ideal connected components, where noises that caused wrong connections are removed and clustering labels can be obtained directly. Finally, we propose an efficient alternating iterative minimization updating strategy, followed by a theoretical proof to prove framework convergence. Comprehensive experiments on five public datasets demonstrate the superiority of the proposed framework.

Where and How to Transfer: Knowledge Aggregation-Induced Transferability Perception for Unsupervised Domain Adaptation.

Unsupervised domain adaptation without accessing expensive annotation processes of target data has achieved remarkable successes in semantic segmentation. However, most existing state-of-the-art methods cannot explore whether semantic representations across domains are transferable or not, which may result in the negative transfer brought by irrelevant knowledge. To tackle this challenge, in this paper, we develop a novel Knowledge Aggregation-induced Transferability Perception (KATP) for unsupervised domain adaptation, which is a pioneering attempt to distinguish transferable or untransferable knowledge across domains. Specifically, the KATP module is designed to quantify which semantic knowledge across domains is transferable, by incorporating transferability information propagation from global category-wise prototypes. Based on KATP, we design a novel KATP Adaptation Network (KATPAN) to determine where and how to transfer. The KATPAN contains a transferable appearance translation module T_A() and a transferable representation augmentation module T_R(), where both modules construct a virtuous circle of performance promotion. T_A() develops a transferability-aware information bottleneck to highlight where to adapt transferable visual characterizations and modality information; T_R() explores how to augment transferable representations while abandoning untransferable information, and promotes the translation performance of T_A() in return. Experiments on several representative datasets and a medical dataset support the state-of-the-art performance of our model.

L3DOC: Lifelong 3D Object Classification.

3D object classification has been widely applied in both academic and industrial scenarios. However, most state-of-the-art algorithms rely on a fixed object classification task set, which cannot tackle the scenario when a new 3D object classification task is coming. Meanwhile, the existing lifelong learning models can easily destroy the learned tasks performance, due to the unordered, large-scale, and irregular 3D geometry data. To address these challenges, we propose a Lifelong 3D Object Classification (i.e., L3DOC) model, which can consecutively learn new 3D object classification tasks via imitating "human learning". More specifically, the core idea of our model is to capture and store the cross-task common knowledge of 3D geometry data in a 3D neural network, named as point-knowledge, through employing layer-wise point-knowledge factorization architecture. Afterwards, a task-relevant knowledge distillation mechanism is employed to connect the current task to previous relevant tasks and effectively prevent catastrophic forgetting. It consists of a point-knowledge distillation module and a transforming-space distillation module, which transfers the accumulated point-knowledge from previous tasks and soft-transfers the compact factorized representations of the transforming-space, respectively. To our best knowledge, the proposed L3DOC algorithm is the first attempt to perform deep learning on 3D object classification tasks in a lifelong learning way. Extensive experiments on several point cloud benchmarks illustrate the superiority of our L3DOC model over the state-of-the-art lifelong learning methods.

What and How: Generalized Lifelong Spectral Clustering via Dual Memory.

Spectral clustering has become one of the most effective clustering algorithms. We in this work explore the problem of spectral clustering in a lifelong learning framework termed as Generalized Lifelong Spectral Clustering (GL 2SC). Different from most current studies, which concentrate on a fixed spectral clustering task set and cannot efficiently incorporate a new clustering task, the goal of our work is to establish a generalized model for new spectral clustering task by What and How to lifelong learn from past tasks. For what to lifelong learn, our GL 2SC framework contains a dual memory mechanism with a deep orthogonal factorization manner: an orthogonal basis memory stores hidden and hierarchical clustering centers among learned tasks, and a feature embedding memory captures deep manifold representation common across multiple related tasks. When a new clustering task arrives, the intuition here for how to lifelong learn is that GL 2SC can transfer intrinsic knowledge from dual memory mechanism to obtain task-specific encoding matrix. Then the encoding matrix can redefine the dual memory over time to provide maximal benefits when learning future tasks. To the end, empirical comparisons on several benchmark datasets show the effectiveness of our GL 2SC, in comparison with several state-of-the-art spectral clustering models.

Nanoparticles: Oral Delivery for Protein and Peptide Drugs.

Protein and peptide drugs have many advantages, such as high bioactivity and specificity, strong solubility, and low toxicity. Therefore, the strategies for improving the bioavailability of protein peptides are reviewed, including chemical modification of nanocarriers, absorption enhancers, and mucous adhesion systems. The status, advantages, and disadvantages of various strategies are systematically analyzed. The systematic and personalized design of various factors affecting the release and absorption of drugs based on nanoparticles is pointed out. It is expected to design a protein peptide oral delivery system that can be applied in the clinic.

[Effect of Iron on the Release of Arsenic in Flooded Paddy Soils].

Amorphous iron oxides in paddy soil are critical adsorbents of arsenic. The flooding period during rice cultivation contributes to the reductive dissolution of these amorphous iron oxides, which releases sorbed arsenic into the paddy soil solution. However, more detailed work should be conducted to evaluate quantitatively arsenic immobilization, release, and transformation regulated by metastable amorphous iron oxides. In previous studies, arsenic in the soil solution phase and solid phase were classified into F1 (exchangeable arsenic), F2 (specifically sorbed arsenic), F3 (amorphous iron oxide bound arsenic), and F4 (crystalline iron oxide bound arsenic), according to a sequential extraction procedure using reagents of increasing dissolution strength. In this study, soil samples were collected from the vicinity of a silver smelting plant in Chenzhou, Hunan Province, and the contribution of different arsenic speciation (F1, F2, F3, and F4) to arsenic release during anaerobic enrichment incubation of paddy soil was investigated. Sample analysis was conducted at the end of the first phase (day 15) and the second phase (day 30). The effects of amorphous iron oxides in paddy soil on migration and transformation of arsenic were discussed. Results showed significant elevation of dissolved Fe(Ⅱ) and arsenic concentration (P<0.05) in enrichment solutions in the second phase compared with that in the first phase. Arsenic released in the soil solution in both phases originated from exchangeable arsenic and specifically sorbed arsenic, as indicated by its significantly positive correlation with F1 and F2 (r=0.73, P<0.05; r=0.657, P<0.05). However, an insignificant positive correlation was found between the arsenic released and F3. Moreover, HCl-extractable Fe(Ⅱ) was significantly and positively correlated with arsenic (r=0.577, P<0.05; r=0.613, P<0.05), while amorphous iron oxides were significantly and negatively correlated with arsenic (r=-0.428, P=0.126; r=-0.564, P<0.05). In conclusion, arsenic in the F1 and F2 fractions acted as the major source of released arsenic. Despite elevated levels of HCl-extractable Fe(Ⅱ) that might result from the slight reductive dissolution of amorphous iron oxide, the significant negative correlation between dissolved arsenic and amorphous iron oxides indicated that metastable amorphous iron oxides in anaerobic paddy soil can generally sorb dissolved arsenic effectively, resulting in lower mobility of arsenic. Increasing the level of amorphous iron oxides in paddy soil is conducive to inactivation of arsenic.

Distribution patterns and major sources of dioxins in soils of the Changsha-Zhuzhou-Xiangtan urban agglomeration, China.

Surface soils from an industrial base, the Changsha-Zhuzhou-Xiangtan urban agglomeration in central China were analyzed for 2378-substituted polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs). The PCDD/F concentrations ranged from 268 to 7510 pg g(-1) dry weight (dw), 72% of which were above the U.S. guideline value (1000 pg g(-1)). It was found that octachlorodibenzo-p-dioxin (OCDD) was the most dominant congener accounting for 78.4-99.3% of the total PCDD/Fs, which was consistent with the PCDD/F profiles reported in other Asian countries. It is recommended that the four major sources of PCDD/Fs in the region can be diesel-fuel vehicles, open straw burning, mass burn-water wall (MB-WW), pentachlorophenate (PCP)/PCP-Na and boilers-hazardous waste incineration. This study is one of the few studies with a focus on the PCDD/F pollution in central China, providing evidences for establishing priorities in reduction of ecological risks posed by PCDD/Fs in central China and elsewhere.

Distribution patterns of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans in sediments of the Xiangjiang River, China.

We investigated the occurrence and distribution patterns of 2,3,7,8-substituted polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) in six sediment samples from the Xiangjiang River, Hunan Province, People's Republic of China. Total concentrations of PCDD/Fs ranged from 876 to 497,759 (mean 160,766) ng/kg dw, the highest of which exceeded that have ever been reported for sediment samples. World Health Organization total toxicity equivalent (WHO-TEQ) concentrations in three out of six samples were significantly higher than the guidance level (21.5 ng WHO-TEQ/kg dw) suggested by Canadian Sediment Quality Guideline. A predominance of octachlorodibenzo-p-dioxin (OCDD) was observed with an average contribution of 90.8% to the total PCDD/F concentrations, while 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD) was the major contributor to the PCDD/F WHO-TEQ concentrations in most of the sites. Such high levels of OCDD and HpCDD may be attributed to the presence of PCP/PCP-Na pollution, although MB-WW, agricultural straw open burning, and boilers-hazardous wastes were also the potential sources of PCDD/Fs. This is the first report for the concentrations and congener profiles of PCDD/Fs in sediment samples from the Xiangtan, Zhuzhou, and Changsha sections of the Xiangjiang River, providing scientific evidence for establishing priorities to reduce ecological risks posed by PCDD/Fs in the rapidly developing areas of Hunan Province and elsewhere.