For underwater applications, the effects of light absorption and scattering result in image degradation. Moreover, the complex and changeable imaging environment makes it difficult to provide a universal enhancement solution to cope with the diversity of water types. In this letter, we present a novel underwater image enhancement (UIE) framework termed SCNet to address the above issues. SCNet is based on normalization schemes across both spatial and channel dimensions with the key idea of learning water type desensitized features. Considering the diversity of degradation is mainly rooted in the strong correlation among pixels, we apply whitening to de-correlates activations across spatial dimensions for each instance in a mini-batch. We also eliminate channel-wise correlation by standardizing and re-injecting the first two moments of the activations across channels. The normalization schemes of spatial and channel dimensions are performed at each scale of the U-Net to obtain multi-scale representations. With such latent encodings, the decoder can easily reconstruct the clean signal, and unaffected by the distortion types caused by the water. Experimental results on two real-world UIE datasets show that the proposed approach can successfully enhance images with diverse water types, and achieves competitive performance in visual quality improvement.
To improve the quality of underwater images, various kinds of underwater image enhancement (UIE) operators have been proposed during the past few years. However, the lack of effective objective evaluation methods limits the further development of UIE techniques. In this paper, we propose a novel rank learning guided no-reference quality assessment method for UIE. Our approach, termed Twice Mixing, is motivated by the observation that a mid-quality image can be generated by mixing a high-quality image with its low-quality version. Typical mixup algorithms linearly interpolate a given pair of input data. However, the human visual system is non-uniformity and non-linear in processing images. Therefore, instead of directly training a deep neural network based on the mixed images and their absolute scores calculated by linear combinations, we propose to train a Siamese Network to learn their quality rankings. Twice Mixing is trained based on an elaborately formulated self-supervision mechanism. Specifically, before each iteration, we randomly generate two mixing ratios which will be employed for both generating virtual images and guiding the network training. In the test phase, a single branch of the network is extracted to predict the quality rankings of different UIE outputs. We conduct extensive experiments on both synthetic and real-world datasets. Experimental results demonstrate that our approach outperforms the previous methods significantly.
Despite deep convolutional neural networks achieved impressive progress in medical image computing and analysis, its paradigm of supervised learning demands a large number of annotations for training to avoid overfitting and achieving promising results. In clinical practices, massive semantic annotations are difficult to acquire in some conditions where specialized biomedical expert knowledge is required, and it is also a common condition where only few annotated classes are available. In this work, we proposed a novel method for few-shot medical image segmentation, which enables a segmentation model to fast generalize to an unseen class with few training images. We construct our few-shot image segmentor using a deep convolutional network trained episodically. Motivated by the spatial consistency and regularity in medical images, we developed an efficient global correlation module to capture the correlation between a support and query image and incorporate it into the deep network called global correlation network. Moreover, we enhance discriminability of deep embedding to encourage clustering of the feature domains of the same class while keep the feature domains of different organs far apart. Ablation Study proved the effectiveness of the proposed global correlation module and discriminative embedding loss. Extensive experiments on anatomical abdomen images on both CT and MRI modalities are performed to demonstrate the state-of-the-art performance of our proposed model.
The effectiveness of existing denoising algorithms typically relies on accurate pre-defined noise statistics or plenty of paired data, which limits their practicality. In this work, we focus on denoising in the more common case where noise statistics and paired data are unavailable. Considering that denoising CNNs require supervision, we develop a new \textbf{adaptive noise imitation (ADANI)} algorithm that can synthesize noisy data from naturally noisy images. To produce realistic noise, a noise generator takes unpaired noisy/clean images as input, where the noisy image is a guide for noise generation. By imposing explicit constraints on the type, level and gradient of noise, the output noise of ADANI will be similar to the guided noise, while keeping the original clean background of the image. Coupling the noisy data output from ADANI with the corresponding ground-truth, a denoising CNN is then trained in a fully-supervised manner. Experiments show that the noisy data produced by ADANI are visually and statistically similar to real ones so that the denoising CNN in our method is competitive to other networks trained with external paired data.
Standard segmentation of medical images based on full-supervised convolutional networks demands accurate dense annotations. Such learning framework is built on laborious manual annotation with restrict demands for expertise, leading to insufficient high-quality labels. To overcome such limitation and exploit massive weakly labeled data, we relaxed the rigid labeling requirement and developed a semi-supervised learning framework based on a teacher-student fashion for organ and lesion segmentation with partial dense-labeled supervision and supplementary loose bounding-box supervision which are easier to acquire. Observing the geometrical relation of an organ and its inner lesions in most cases, we propose a hierarchical organ-to-lesion (O2L) attention module in a teacher segmentor to produce pseudo-labels. Then a student segmentor is trained with combinations of manual-labeled and pseudo-labeled annotations. We further proposed a localization branch realized via an aggregation of high-level features in a deep decoder to predict locations of organ and lesion, which enriches student segmentor with precise localization information. We validated each design in our model on LiTS challenge datasets by ablation study and showed its state-of-the-art performance compared with recent methods. We show our model is robust to the quality of bounding box and achieves comparable performance compared with full-supervised learning methods.
Retinal artery/vein (A/V) classification plays a critical role in the clinical biomarker study of how various systemic and cardiovascular diseases affect the retinal vessels. Conventional methods of automated A/V classification are generally complicated and heavily depend on the accurate vessel segmentation. In this paper, we propose a multi-task deep neural network with spatial activation mechanism that is able to segment full retinal vessel, artery and vein simultaneously, without the pre-requirement of vessel segmentation. The input module of the network integrates the domain knowledge of widely used retinal preprocessing and vessel enhancement techniques. We specially customize the output block of the network with a spatial activation mechanism, which takes advantage of a relatively easier task of vessel segmentation and exploits it to boost the performance of A/V classification. In addition, deep supervision is introduced to the network to assist the low level layers to extract more semantic information. The proposed network achieves pixel-wise accuracy of 95.70% for vessel segmentation, and A/V classification accuracy of 94.50%, which is the state-of-the-art performance for both tasks on the AV-DRIVE dataset. Furthermore, we have also tested the model performance on INSPIRE-AVR dataset, which achieves a skeletal A/V classification accuracy of 91.6%.
Accurate computing, analysis and modeling of the ventricles and myocardium from medical images are important, especially in the diagnosis and treatment management for patients suffering from myocardial infarction (MI). Late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) provides an important protocol to visualize MI. However, automated segmentation of LGE CMR is still challenging, due to the indistinguishable boundaries, heterogeneous intensity distribution and complex enhancement patterns of pathological myocardium from LGE CMR. Furthermore, compared with the other sequences LGE CMR images with gold standard labels are particularly limited, which represents another obstacle for developing novel algorithms for automatic segmentation of LGE CMR. This paper presents the selective results from the Multi-Sequence Cardiac MR (MS-CMR) Segmentation challenge, in conjunction with MICCAI 2019. The challenge offered a data set of paired MS-CMR images, including auxiliary CMR sequences as well as LGE CMR, from 45 patients who underwent cardiomyopathy. It was aimed to develop new algorithms, as well as benchmark existing ones for LGE CMR segmentation and compare them objectively. In addition, the paired MS-CMR images could enable algorithms to combine the complementary information from the other sequences for the segmentation of LGE CMR. Nine representative works were selected for evaluation and comparisons, among which three methods are unsupervised methods and the other six are supervised. The results showed that the average performance of the nine methods was comparable to the inter-observer variations. The success of these methods was mainly attributed to the inclusion of the auxiliary sequences from the MS-CMR images, which provide important label information for the training of deep neural networks.
Recent advances in adaptive object detection have achieved compelling results in virtue of adversarial feature adaptation to mitigate the distributional shifts along the detection pipeline. Whilst adversarial adaptation significantly enhances the transferability of feature representations, the feature discriminability of object detectors remains less investigated. Moreover, transferability and discriminability may come at a contradiction in adversarial adaptation given the complex combinations of objects and the differentiated scene layouts between domains. In this paper, we propose a Hierarchical Transferability Calibration Network (HTCN) that hierarchically (local-region/image/instance) calibrates the transferability of feature representations for harmonizing transferability and discriminability. The proposed model consists of three components: (1) Importance Weighted Adversarial Training with input Interpolation (IWAT-I), which strengthens the global discriminability by re-weighting the interpolated image-level features; (2) Context-aware Instance-Level Alignment (CILA) module, which enhances the local discriminability by capturing the underlying complementary effect between the instance-level feature and the global context information for the instance-level feature alignment; (3) local feature masks that calibrate the local transferability to provide semantic guidance for the following discriminative pattern alignment. Experimental results show that HTCN significantly outperforms the state-of-the-art methods on benchmark datasets.
Referring Expression Comprehension (REC) is an emerging research spot in computer vision, which refers to detecting the target region in an image given an text description. Most existing REC methods follow a multi-stage pipeline, which are computationally expensive and greatly limit the application of REC. In this paper, we propose a one-stage model towards real-time REC, termed Real-time Global Inference Network (RealGIN). RealGIN addresses the diversity and complexity issues in REC with two innovative designs: the Adaptive Feature Selection (AFS) and the Global Attentive ReAsoNing unit (GARAN). AFS adaptively fuses features at different semantic levels to handle the varying content of expressions. GARAN uses the textual feature as a pivot to collect expression-related visual information from all regions, and thenselectively diffuse such information back to all regions, which provides sufficient context for modeling the complex linguistic conditions in expressions. On five benchmark datasets, i.e., RefCOCO, RefCOCO+, RefCOCOg, ReferIt and Flickr30k, the proposed RealGIN outperforms most prior works and achieves very competitive performances against the most advanced method, i.e., MAttNet. Most importantly, under the same hardware, RealGIN can boost the processing speed by about 10 times over the existing methods.
The performance of a deep neural network is highly dependent on its training, and finding better local optimal solutions is the goal of many optimization algorithms. However, existing optimization algorithms show a preference for descent paths that converge slowly and do not seek to avoid bad local optima. In this work, we propose Learning Rate Dropout (LRD), a simple gradient descent technique for training related to coordinate descent. LRD empirically aids the optimizer to actively explore in the parameter space by randomly setting some learning rates to zero; at each iteration, only parameters whose learning rate is not 0 are updated. As the learning rate of different parameters is dropped, the optimizer will sample a new loss descent path for the current update. The uncertainty of the descent path helps the model avoid saddle points and bad local minima. Experiments show that LRD is surprisingly effective in accelerating training while preventing overfitting.