Santiagoelmore9685
The success of deep convolutional networks (ConvNets) generally relies on a massive amount of well-labeled data, which is labor-intensive and time-consuming to collect and annotate in many scenarios. To eliminate such limitation, self-supervised learning (SSL) is recently proposed. Specifically, by solving a pre-designed proxy task, SSL is capable of capturing general-purpose features without requiring human supervision. Existing efforts focus obsessively on designing a particular proxy task but ignore the semanticity of samples that are advantageous to downstream tasks, resulting in the inherent limitation that the learned features are specific to the proxy task, namely the proxy task-specificity of features. In this work, to improve the generalizability of features learned by existing SSL methods, we present a novel self-supervised framework SSL++ to incorporate the proxy task-independent semanticity of samples into the representation learning process. MEK inhibitor review Technically, SSL++ aims to leverage the complementarity, between the low-level generic features learned by a proxy task and the high-level semantic features newly learned by the generated semantic pseudo-labels, to mitigate the task-specificity and improve the generalizability of features. Extensive experiments show that SSL++ performs favorably against the state-of-the-art approaches on the established and latest SSL benchmarks.This work proposes the neural reference synthesis (NRS) to generate high-fidelity reference block for motion estimation and motion compensation (MEMC) in inter frame coding. The NRS is comprised of two submodules one for reconstruction enhancement and the other for reference generation. Although numerous methods have been developed in the past for these two submodules using either handcrafted rules or deep convolutional neural network (CNN) models, they basically deal with them separately, resulting in limited coding gains. By contrast, the NRS proposes to optimize them collaboratively. It first develops two CNN-based models, namely EnhNet and GenNet. The EnhNet only uses spatial correlations within the current frame for reconstruction enhancement and the GenNet is then augmented by further aggregating temporal correlations across multiple frames for reference synthesis. However, a direct concatenation of EnhNet and GenNet without considering the complex temporal reference dependency across inter frames would implicitly induce iterative CNN processing and cause the data overfitting problem, leading to visually-disturbing artifacts and oversmoothed pixels. To tackle this problem, the NRS applies a new training strategy to coordinate the EnhNet and GenNet for more robust and generalizable models, and also devises a lightweight multi-level R-D (rate-distortion) selection policy for the encoder to adaptively choose reference blocks generated from the proposed NRS model or conventional coding process. Our NRS not only offers state-of-the-art coding gains, e.g., >10% BD-Rate (Bjøntegaard Delta Rate) reduction against the High Efficiency Video Coding (HEVC) anchor for a variety of common test video sequences encoded at a wide bit range in both low-delay and random access settings, but also greatly reduces the complexity relative to existing learning-based methods by utilizing more lightweight DNNs. All models are made publicly accessible at https//github.com/IVC-Projects/NRS for reproducible research.The thrombolysis potential of low-boiling-point (-2 °C) perfluorocarbon phase-change nanodroplets (NDs) has previously been demonstrated on aged clots, and we hypothesized that this efficacy would extend to retracted clots. We tested this hypothesis by comparing sonothrombolysis of both unretracted and retracted clots using ND-mediated ultrasound (US+ND) and microbubble-mediated ultrasound (US+MB), respectively. Assessment data included clot mass reduction, cavitation detection, and cavitation cloud imaging in vitro. Acoustic parameters included a 7.9-MPa peak negative pressure and 180-cycle bursts with 5-Hz repetition (the corresponding duty cycle and time-averaged intensity of 0.09% and 1.87 W/cm2, respectively) based on prior studies. With these parameters, we observed a significantly reduced efficacy of US+MB in the retracted versus unretracted model (the averaged mass reduction rate from 1.83%/min to 0.54%/min). Unlike US+MB, US+ND exhibited less reduction of efficacy in the retracted model (from 2.15%/min to 1.04%/min on average). The cavitation detection results correlate with the sonothrombolysis efficacy results showing that both stable and inertial cavitation generated in a retracted clot by US+ND is higher than that by US+MB. We observed that ND-mediated cavitation shows a tendency to occur inside a clot, whereas MB-mediated cavitation occurs near the surface of a retracted clot, and this difference is more significant with retracted clots compared to unretracted clots. We conclude that ND-mediated sonothrombolysis outperforms MB-mediated therapy regardless of clot retraction, and this advantage of ND-mediated cavitation is emphasized for retracted clots. The primary mechanisms are hypothesized to be sustained cavitation level and cavitation clouds in the proximity of a retracted clot by US+ND.Deformable registration is fundamental to longitudinal and population-based image analyses. However, it is challenging to precisely align longitudinal infant brain MR images of the same subject, as well as cross-sectional infant brain MR images of different subjects, due to fast brain development during infancy. In this paper, we propose a recurrently usable deep neural network for the registration of infant brain MR images. There are three main highlights of our proposed method. (i)We use brain tissue segmentation maps for registration, instead of intensity images, to tackle the issue of rapid contrast changes of brain tissues during the first year of life. (ii) A single registration network is trained in a one-shot manner, and then recurrently applied in inference for multiple times, such that the complex deformation field can be recovered incrementally. (iii) We also propose both the adaptive smoothing layer and the tissue-aware anti-folding constraint into the registration network to ensure the physiological plausibility of estimated deformations without degrading the registration accuracy. Experimental results, in comparison to the state-of-the-art registration methods, indicate that our proposed method achieves the highest registration accuracy while still preserving the smoothness of the deformation field. The implementation of our proposed registration network is available online.Spectral clustering (SC) algorithms have been successful in discovering meaningful patterns since they can group arbitrarily shaped data structure. Traditional SC approaches typically consist of two sequential stages, i.e., performing spectral decomposition of an affinity matrix and then rounding the relaxed continuous clustering result into a discrete indicator matrix. However, such a two-stage process could make the obtained discrete indicator matrix severely deviate from the ground true one. This is because the former step is not devoted to achieving an optimal clustering result. To alleviate this issue, this paper presents a general joint framework to simultaneously learn the optimal continuous and binary indicator matrices for multi-view clustering, which also has the ability to tackle the conventional single-view case. Specially, we provide a theoretical proof for the proposed method. Furthermore, an effective alternate updating algorithm is developed to optimize the corresponding complex objective. A number of empirical results on different benchmark datasets demonstrate the proposed method outperforms several state-of-the-art in terms of seven clustering metrics.Recently, hyperbolic deep neural networks (HDNNs) have been gaining momentum as the deep representations in the hyperbolic space provide high fidelity embeddings with few dimensions, especially for data possessing hierarchical structure. Such a hyperbolic neural architecture is quickly extended to many different scientific fields, including natural language processing, single-cell RNA-sequence analysis, graph embedding, financial analysis, and computer vision. The promising results demonstrate its superior capability, significant compactness of the model, and a substantially better physical interpretability than its counterpart in the Euclidean space. To stimulate future research, this paper presents a coherent and a comprehensive review of the literature around the neural components in the construction of HDNN, as well as the generalization of the leading deep approaches to the hyperbolic space. It also presents current applications of various tasks, together with insightful observations and identifying open questions and promising future directions.Monocular 3D object detection is an important task in autonomous driving. It can be easily intractable where there exists ego-car pose change w.r.t. ground plane. This is common due to the slight fluctuation of road smoothness and slope. Due to the lack of insight in industrial application, existing methods on open datasets neglect camera pose information, which inevitably results in the detector being susceptible to camera extrinsic parameters. The perturbation of objects is very popular in most autonomous driving cases for industrial products. To this end, we propose a novel method to capture camera pose to formulate the detector free from extrinsic perturbation. Specifically, the proposed framework predicts camera extrinsic parameters by detecting vanishing point and horizon change. A converter is designed to rectify perturbative features in the latent space. By doing so, our 3D detector works independent of the extrinsic parameter variations, and produces accurate results in realistic cases, e.g., potholed and uneven roads, where almost all existing monocular detectors fail to handle. Experiments demonstrate our method yields best performance compared with the other state-of-the-arts by a large margin on both KITTI 3D and nuScenes datasets.
Ultrasound Localization Microscopy (ULM) provides images of the microcirculation in-depth in living tissue. However, its implementation in two-dimension is limited by the elevation projection and tedious plane-by-plane acquisition. Volumetric ULM alleviates these issues and can map the vasculature of entire organs in one acquisition with isotropic resolution. However, its optimal implementation requires many independent acquisition channels, leading to complex custom hardware.
In this article, we implemented volumetric ultrasound imaging with a multiplexed 32 x 32 probe driven by a single commercial ultrasound scanner. We propose and compare three different sub-aperture multiplexing combinations for localization microscopy in silico and in vitro with a flow of microbubbles in a canal. Finally, we evaluate the approach for micro-angiography of the rat brain.The "light" combination allows a higher maximal volume rate than the "full" combination while maintaining the field of view and resolution.
In the rat brain, 100,000 volumes were acquired within 7 min with a dedicated ultrasound sequence and revealed vessels down to 31 m in diameter with flows from 4.3 mm/s to 28.4 mm/s.
This work demonstrates the ability to perform a complete angiography with unprecedented resolution in the living rats brain with a simple and light setup through the intact skull.
We foresee that it might contribute to democratize 3D ULM for both preclinical and clinical studies.
We foresee that it might contribute to democratize 3D ULM for both preclinical and clinical studies.