The uncertainty of water cloud feedback on radiative forcing is one of the largest obstacles to producing confident projections of the global climate. Sufficient measurements of water clouds are crucial to addressing this issue. However, existing techniques based on remote sensing or in situ instruments face limitations in data capacity attributed to the short lifetime, high temporal variability, and complex vertical structure of water clouds. In this study, taking advantage of a dual-field-of-view (dual-FOV) high-spectral-resolution lidar (HSRL), we developed a novel algorithm to obtain diurnal simultaneous profiles of water cloud optical and microphysical properties with high temporal-spatial resolution. This technique does not rely on the widely used subadiabatic assumption about the vertical structure of water clouds. The retrieval algorithm, validated by simulations and cloud radar measurements, was applied to field experiment data collected at the Beijing and Hangzhou sites in China. The relationship functions between water cloud properties are presented to enhance our understanding of the underlying processes. Furthermore, the vertical distributions of retrieved properties are compared to the subadiabatic assumption. The dual-FOV HSRL technique enables comprehensive observations, enhancing our understanding of water clouds and providing significant insights into the interactions among clouds, aerosols, precipitation, and radiation.

This paper investigates how to effectively mine contextual information among sequential images and jointly model them in medical imaging tasks. Different from state-of-the-art methods that model sequential correlations via point-wise token encoding, this paper develops a novel hierarchical pattern-aware tokenization strategy. It handles distinct visual patterns independently and hierarchically, which not only ensures the full flexibility of attention aggregation under different pattern representations but also preserves both local and global information simultaneously. Based on this strategy, we propose a Pattern-Aware Transformer (PATrans) featuring a global-local dual-path pattern-aware cross-attention mechanism to achieve hierarchical pattern matching and propagation among sequential images. Furthermore, PATrans is plug-and-play and can be seamlessly integrated into various backbone networks for diverse downstream sequence modeling tasks. We demonstrate its general application paradigm across four domains and five benchmarks in video object detection and 3D volumetric semantic segmentation tasks, respectively. Impressively, PATrans sets new state-of-the-art across all these benchmarks, i.e., CVC-Video (92.3% detection F1), ASU-Mayo (99.1% localization F1), Lung Tumor (78.59% DSC), Nasopharynx Tumor (75.50% DSC), and Kidney Tumor (87.53% DSC). Codes and models are available at https://github.com/GGaoxiang/PATrans.

Facial image manipulation has achieved great progress in recent years. However, previous methods either operate on a predefined set of face attributes or leave users little freedom to interactively manipulate images. To overcome these drawbacks, we propose a novel framework termed MaskGAN, enabling diverse and interactive face manipulation. Our key insight is that semantic masks serve as a suitable intermediate representation for flexible face manipulation with fidelity preservation. MaskGAN has two main components: 1) Dense Mapping Network (DMN) and 2) Editing Behavior Simulated Training (EBST). Specifically, DMN learns style mapping between a free-form user modified mask and a target image, enabling diverse generation results. EBST models the user editing behavior on the source mask, making the overall framework more robust to various manipulated inputs. Specifically, it introduces dual-editing consistency as the auxiliary supervision signal. To facilitate extensive studies, we construct a large-scale high-resolution face dataset with fine-grained mask annotations named CelebAMask-HQ. MaskGAN is comprehensively evaluated on two challenging tasks: attribute transfer and style copy, demonstrating superior performance over other state-of-the-art methods. The code, models, and dataset are available at https://github.com/switchablenorms/CelebAMask-HQ.

Normalization techniques are important in different advanced neural networks and different tasks. This work investigates a novel dynamic learning-to-normalize (L2N) problem by proposing Exemplar Normalization (EN), which is able to learn different normalization methods for different convolutional layers and image samples of a deep network. EN significantly improves the flexibility of the recently proposed switchable normalization (SN), which solves a static L2N problem by linearly combining several normalizers in each normalization layer (the combination is the same for all samples). Instead of directly employing a multi-layer perceptron (MLP) to learn data-dependent parameters as conditional batch normalization (cBN) did, the internal architecture of EN is carefully designed to stabilize its optimization, leading to many appealing benefits. (1) EN enables different convolutional layers, image samples, categories, benchmarks, and tasks to use different normalization methods, shedding light on analyzing them in a holistic view. (2) EN is effective for various network architectures and tasks. (3) It could replace any normalization layers in a deep network and still produce stable model training. Extensive experiments demonstrate the effectiveness of EN in a wide spectrum of tasks including image recognition, noisy label learning, and semantic segmentation. For example, by replacing BN in the ordinary ResNet50, improvement produced by EN is 300% more than that of SN on both ImageNet and the noisy WebVision dataset. The codes and models will be released.

This work presents Dynamic Normalization (DN), which is able to learn arbitrary normalization operations for different convolutional layers in a deep ConvNet. Unlike existing normalization approaches that predefined computations of the statistics (mean and variance), DN learns to estimate them. DN has several appealing benefits. First, it adapts to various networks, tasks, and batch sizes. Second, it can be easily implemented and trained in a differentiable end-to-end manner with merely small number of parameters. Third, its matrix formulation represents a wide range of normalization methods, shedding light on analyzing them theoretically. Extensive studies show that DN outperforms its counterparts in CIFAR10 and ImageNet.

Scheduling the training procedure of deep neural networks (DNNs) such as tuning the learning rates is crucial to the successes of deep learning. Previous strategies such as piecewise and exponential learning rate schedulers have different arguments (hyper-parameters) that need to be tuned manually. With the expanding of data scale and model computation, searching for these arguments spends lots of empirical efforts. To address this issue, this work proposes policy schedular that determines the arguments of learning rate (lr) by reinforcement learning, significantly reducing costs to tune them. The policy schedular possesses several appealing benefits. First, instead of manually defining the values of initial lr and ultimate lr, it autonomously determines these values in training. Second, rather than using predefined functions to update lr, it adaptively oscillates lr by monitoring learning curves without human intervention. Third, it is able to select lr for each block or layer of a DNN. Experiments show that the DNNs trained with policy schedular achieve superior performances, outperforming previous work on various tasks and benchmarks such as ImageNet, COCO, and learning-to-learn.

Pretraining visual features by image classification on ImageNet is an indispensable step towards many advanced perception systems in the last decade. ImageNet is the most prevalent database for supervised pretraining of image features. Unlike ImageNet assuming that the visual concepts are static and independent with each other, this work presents a neverending learning platform, termed CUImage, which learns visual representation on a knowledge graph of billions of images, whose data scale is several orders of magnitude larger than ImageNet. A novel dynamic graph convolutional network (GCN) is proposed to learn visual concepts. Once the new data are presented, the GCN is updated dynamically where new concepts can be discovered or existing concepts can be merged. This is enabled by three main components in CUImage, including Data Dispersion (DD), Data Management and Mining (DMM), and Data Evaluation (DE). These three components are built on top of a computer cluster with thousands of GPU/CPU cores and a parallel storage of petabytes. So far, CUImage has processed and managed more than 2 million visual concepts of 2 billion images. To evaluate the learned representation, we transfer the pretrained features to several challenging benchmarks such as image recognition on ImageNet and object detection in MS-COCO. We achieve state-of-the-art results, significantly surpassing the systems that used ImageNet for pretraining. The codes, data, and models will be released.

Boundary incompleteness raises great challenges to automatic prostate segmentation in ultrasound images. Shape prior can provide strong guidance in estimating the missing boundary, but traditional shape models often suffer from hand-crafted descriptors and local information loss in the fitting procedure. In this paper, we attempt to address those issues with a novel framework. The proposed framework can seamlessly integrate feature extraction and shape prior exploring, and estimate the complete boundary with a sequential manner. Our framework is composed of three key modules. Firstly, we serialize the static 2D prostate ultrasound images into dynamic sequences and then predict prostate shapes by sequentially exploring shape priors. Intuitively, we propose to learn the shape prior with the biologically plausible Recurrent Neural Networks (RNNs). This module is corroborated to be effective in dealing with the boundary incompleteness. Secondly, to alleviate the bias caused by different serialization manners, we propose a multi-view fusion strategy to merge shape predictions obtained from different perspectives. Thirdly, we further implant the RNN core into a multiscale Auto-Context scheme to successively refine the details of the shape prediction map. With extensive validation on challenging prostate ultrasound images, our framework bridges severe boundary incompleteness and achieves the best performance in prostate boundary delineation when compared with several advanced methods. Additionally, our approach is general and can be extended to other medical image segmentation tasks, where boundary incompleteness is one of the main challenges.

Ultrasound (US) imaging is a widely used screening tool for obstetric examination and diagnosis. Accurate acquisition of fetal standard planes with key anatomical structures is very crucial for substantial biometric measurement and diagnosis. However, the standard plane acquisition is a labor-intensive task and requires operator equipped with a thorough knowledge of fetal anatomy. Therefore, automatic approaches are highly demanded in clinical practice to alleviate the workload and boost the examination efficiency. The automatic detection of standard planes from US videos remains a challenging problem due to the high intraclass and low interclass variations of standard planes, and the relatively low image quality. Unlike previous studies which were specifically designed for individual anatomical standard planes, respectively, we present a general framework for the automatic identification of different standard planes from US videos. Distinct from conventional way that devises hand-crafted visual features for detection, our framework explores in- and between-plane feature learning with a novel composite framework of the convolutional and recurrent neural networks. To further address the issue of limited training data, a multitask learning framework is implemented to exploit common knowledge across detection tasks of distinctive standard planes for the augmentation of feature learning. Extensive experiments have been conducted on hundreds of US fetus videos to corroborate the better efficacy of the proposed framework on the difficult standard plane detection problem.

The quality of ultrasound (US) images for the obstetric examination is crucial for accurate biometric measurement. However, manual quality control is a labor intensive process and often impractical in a clinical setting. To improve the efficiency of examination and alleviate the measurement error caused by improper US scanning operation and slice selection, a computerized fetal US image quality assessment (FUIQA) scheme is proposed to assist the implementation of US image quality control in the clinical obstetric examination. The proposed FUIQA is realized with two deep convolutional neural network models, which are denoted as L-CNN and C-CNN, respectively. The L-CNN aims to find the region of interest (ROI) of the fetal abdominal region in the US image. Based on the ROI found by the L-CNN, the C-CNN evaluates the image quality by assessing the goodness of depiction for the key structures of stomach bubble and umbilical vein. To further boost the performance of the L-CNN, we augment the input sources of the neural network with the local phase features along with the original US data. It will be shown that the heterogeneous input sources will help to improve the performance of the L-CNN. The performance of the proposed FUIQA is compared with the subjective image quality evaluation results from three medical doctors. With comprehensive experiments, it will be illustrated that the computerized assessment with our FUIQA scheme can be comparable to the subjective ratings from medical doctors.