Search

Stitched fluorescence microscope images inevitably exist in various types of stripes or artifacts caused by uncertain factors such as optical devices or specimens, which severely affects the image quality and downstream quantitative analysis. Here, we present a deep learning-based Stripe Self-Correction method, so-called SSCOR. Specifically, we propose a proximity sampling scheme and adversarial reciprocal self-training paradigm that enable SSCOR to utilize stripe-free patches sampled from the stitched microscope image itself to correct their adjacent stripe patches. Comparing to off-the-shelf approaches, SSCOR can not only adaptively correct non-uniform, oblique, and grid stripes, but also remove scanning, bubble, and out-of-focus artifacts, achieving the state-of-the-art performance across different imaging conditions and modalities. Moreover, SSCOR does not require any physical parameter estimation, patch-wise manual annotation, or raw stitched information in the correction process. This provides an intelligent prior-free image restoration solution for microscopists or even microscope companies, thus ensuring more precise biomedical applications for researchers. Image stitching in fluorescence microscopy can be a hindrance to image quality and to downstream quantitative analyses. Here, the authors propose a deep learning-based stripe self-correction method that corrects diverse stripes and artifacts for stitched microscopic images. [ABSTRACT FROM AUTHOR]

Generic multiple object tracking aims to recover the trajectories for generic moving objects of the same category. This task relies on the ability of effectively extracting representative features of the target objects. To this end, we propose a novel prototype learning based model, PLGMOT, that can explore the template features of an exemplar object and extend to more objects to acquire their prototype. Their prototype features can be continuously updated during the video, in favor of generalization to all the target objects with different appearances. More importantly, on the public benchmark GMOT-40, our method achieves more than 14% advantage over the state-of-the-art methods, with less than 0.5% of the training data that is not even completely annotated in the form of bounding boxes, thanks to our proposed point-to-box label refinement training algorithm and hierarchical motion-aware association algorithm. • A prototype learning based generic multi-object detector. • A point-to-box label refinement training algorithm. • A hierarchical motion-aware association algorithm for tracking. • Extensive experiments demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Crowd counting is challenging due to unconstrained imaging factors, e.g., background clutters, non-uniform distribution of people, large scale and perspective variations. Dealing with these problems using deep neural networks requires rich prior knowledge and multi-scale contextual representations. In this paper, we propose a Cross-stage Refinement Network (CRNet) that can refine predicted density maps progressively based on hierarchical multi-level density priors. In particular, CRNet is composed of several fully convolutional networks. They are stacked together recursively with the previous output as the next input, and each of them serves to utilize previous density output to gradually correct prediction errors of crowd areas and refine the predicted density maps at different stages. Cross-stage multi-level density priors are further exploited in our recurrent framework by the cross-stage skip layers based on ConvLSTM. To cope with different challenges of unconstrained crowd scenes, we explore different crowd-specific data augmentation methods to mimic real-world scenarios and enrich crowd feature representations from different aspects. Extensive experiments show the proposed method achieves superior performances against state-of-the-art methods on four widely-used challenging benchmarks in terms of counting accuracy and density map quality. Code and models are available at this https://github.com/lytgftyf/Crowd-Counting-via-Cross-stage-Refinement-Networks. [ABSTRACT FROM AUTHOR]

Developing a safe and efficient collision-avoidance policy for multiple robots is challenging in the decentralized scenarios where each robot generates its paths with limited observation of other robots' states and intentions. Prior distributed multi-robot collision-avoidance systems often require frequent inter-robot communication or agent-level features to plan a local collision-free action, which is not robust and computationally prohibitive. In addition, the performance of these methods is not comparable with their centralized counterparts in practice. In this article, we present a decentralized sensor-level collision-avoidance policy for multi-robot systems, which shows promising results in practical applications. In particular, our policy directly maps raw sensor measurements to an agent's steering commands in terms of the movement velocity. As a first step toward reducing the performance gap between decentralized and centralized methods, we present a multi-scenario multi-stage training framework to learn an optimal policy. The policy is trained over a large number of robots in rich, complex environments simultaneously using a policy-gradient-based reinforcement-learning algorithm. The learning algorithm is also integrated into a hybrid control framework to further improve the policy's robustness and effectiveness. We validate the learned sensor-level collision-3avoidance policy in a variety of simulated and real-world scenarios with thorough performance evaluations for large-scale multi-robot systems. The generalization of the learned policy is verified in a set of unseen scenarios including the navigation of a group of heterogeneous robots and a large-scale scenario with 100 robots. Although the policy is trained using simulation data only, we have successfully deployed it on physical robots with shapes and dynamics characteristics that are different from the simulated agents, in order to demonstrate the controller's robustness against the simulation-to-real modeling error. Finally, we show that the collision-avoidance policy learned from multi-robot navigation tasks provides an excellent solution for safe and effective autonomous navigation for a single robot working in a dense real human crowd. Our learned policy enables a robot to make effective progress in a crowd without getting stuck. More importantly, the policy has been successfully deployed on different types of physical robot platforms without tedious parameter tuning. Videos are available at https://sites.google.com/view/hybridmrca. [ABSTRACT FROM AUTHOR]

Physical behaviors (e.g., physical activity and sedentary behavior) have been the focus among many researchers in the biomedical and behavioral science fields. The recent shift from hip- to wrist-worn accelerometers in these fields has signaled the need to develop novel approaches to process raw acceleration data of physical activity and sedentary behavior. However, there is currently no consensus regarding the best practices for analyzing wrist-worn accelerometer data to accurately predict individuals' energy expenditure and the times spent in different intensities of free-living physical activity and sedentary behavior. To this end, accurately analyzing and interpreting wrist-worn accelerometer data has become a major challenge facing many clinicians and researchers. In response, this paper attempts to review different methodologies for analyzing wrist-worn accelerometer data and offer cutting edge, yet appropriate analysis plans for wrist-worn accelerometer data in the assessment of physical behavior. In this paper, we first discuss the fundamentals of wrist-worn accelerometer data, followed by various methods of processing these data (e.g., cut points, steps per minute, machine learning), and then we discuss the opportunities, challenges, and directions for future studies in this area of inquiry. This is the most comprehensive review paper to date regarding the analysis and interpretation of free-living physical activity data derived from wrist-worn accelerometers, aiming to help establish a blueprint for processing wrist-derived accelerometer data. [ABSTRACT FROM AUTHOR]

Domain Generalization (DG) aims to generalize a model trained on multiple source domains to an unseen target domain. The source domains always require precise annotations, which can be cumbersome or even infeasible to obtain in practice due to the vast amount of data involved. Web data, namely web-crawled images, offers an opportunity to access large amounts of unlabeled images with rich style information, which can be leveraged to improve DG. From this perspective, we introduce a novel paradigm of DG, termed as Semi-Supervised Domain Generalization (SSDG) , to explore how the labeled and unlabeled source domains can interact, and establish two settings, including the close-set and open-set SSDG. The close-set SSDG is based on existing public DG datasets, while the open-set SSDG, built on the newly-collected web-crawled datasets, presents a novel yet realistic challenge that pushes the limits of current technologies. A natural approach of SSDG is to transfer knowledge from labeled data to unlabeled data via pseudo labeling, and train the model on both labeled and pseudo-labeled data for generalization. Since there are conflicting goals between domain-oriented pseudo labeling and out-of-domain generalization, we develop a pseudo labeling phase and a generalization phase independently for SSDG. Unfortunately, due to the large domain gap, the pseudo labels provided in the pseudo labeling phase inevitably contain noise, which has negative affect on the subsequent generalization phase. Therefore, to improve the quality of pseudo labels and further enhance generalizability, we propose a cyclic learning framework to encourage a positive feedback between these two phases, utilizing an evolving intermediate domain that bridges the labeled and unlabeled domains in a curriculum learning manner. Extensive experiments are conducted to validate the effectiveness of our method. It is worth highlighting that web-crawled images can promote domain generalization as demonstrated by the experimental results. • Developing the close-set and open-set semi-supervised domain generalization (SSDG). • Constructing two web-crawled datasets are for the open-set SSDG. • Establishing strong baselines for SSDG using existing deep learning methods. • Proposing a framework with evolving intermediate domain for SSDG. • The framework outperforms the strong baselines in both close-set and open-set SSDG. [ABSTRACT FROM AUTHOR]

• Deep compression of these deterministic models has been proposed in last few years to reduce the number of connections and nodes, while remaining the classification accuracy of these models. This paper is the first attempt to combine deep probabilistic graphical networks (PGNs) and deep compression techniques together to derive sparse versions of the deep probabilistic models. • The developed pruning approach is a layer-by-layer and pre-pruning one. That is, the final PGNs grow in a layer-by-layer and compressed manner. It is very efficient in redundancy reduction. The whole pruning only takes a few seconds to conduct, while other complex methods take much more time to select sparse architectures or conduct model compression, such as regularization and low-rank factorization approaches. These model compression techniques are more time-consuming compared with pruning due to the complexity of regularization and factorization. • The correctness and efficiency of the proposed compression approach are evaluated by a number of PGNs with different datasets. It can be used to compress trained PGNs to achieve a lightweight network which can be embedded in on-chip memory of small mobile devices. It is also flexible to implement and can be generalized to other deep PGNs. Probabilistic Graphical Models (PGMs) are important and active research areas in machine learning and artificial intelligence. The well-known representatives of PGMs include Restricted Boltzmann Machines (RBMs), Deep Belief Networks (DBNs), Deep Boltzmann Machines (DBMs), and their variants. These PGMs open a new dimension of machine learning and directly lead to the revival of deep learning since 2006. Compared with deep deterministic models, such as deep convolutional neural networks (CNNs), PGMs are more flexible, robust and powerful for both unsupervised and supervised learning problems. As the generative models on the one hand, PGMs have the good capability of generating samples after extracting informative features of input data. On the other hand, they also can provide a good initialization of hierarchical structures for supervised learning. The PGMs with deep structures are called deep Probabilistic Graphical Networks (PGNs) due to their hierarchical structures and close relationships with deep neural networks. There are a large number of CNN-based deep deterministic networks proposed for computer vision tasks. However, these deep deterministic models are shown to possess extremely large number of model parameters, and have serious redundancy problems. Therefore, deep compression of these deterministic models have been proposed in the last few years to reduce the number of connections and nodes, while preserving the classification accuracy. This paper is the first attempt to combine deep PGNs and deep compression techniques together to derive sparse versions of the deep probabilistic models. We firstly examine that whether deep PGNs have serious problem of parameter redundancy, and compress them if necessary. From the experimental results on MNIST, Fashion-MNIST and CIFAR-10 dataset, deep PGNs can be compressed by at least 50% of their parameters through a deep pruning and retraining approach, while keeping their capabilities in both generative and discriminative tasks. [ABSTRACT FROM AUTHOR]