Alpha Channel Research Articles

Semi-Supervised Portrait Matting via the Collaboration of Teacher–Student Network and Adaptive Strategies

In the portrait matting domain, existing methods rely entirely on annotated images for learning. However, delicate manual annotations are time-consuming and there are few detailed datasets available. To reduce complete dependency on labeled datasets, we design a semi-supervised network (ASSN) with two kinds of innovative adaptive strategies for portrait matting. Three pivotal sub-modules are embedded in our architecture, including a static teacher network (S-TN), a static student network (S-SN), and an adaptive student network (A-SN). S-TN and S-SN are modules that need to be trained with a small number of high-quality labeled datasets. Moreover, A-SN and S-SN share the same module parameters. When processing unlabeled datasets, A-SN adopts the adaptive strategies designed by us to discard the dependence on labeled datasets. The adaptive strategies include: (i) An auxiliary adaption: The teacher network with complicated design not only provides alpha mattes for the adaptive student network but also transmits rough segmentation results and edge graphs as optimization reference standards. (ii) A self-adjusting adaption: The adaptive network can make self-supervised to the characteristics of different layers. In addition, we have produced a finely annotated dataset for scholars in the field. Compared with existing datasets, our dataset complements the following two types of data neglected in previous datasets: (i) Images taken by multiple people. (ii) Images under low light conditions.

Improving the feasibility of economical proton-boron-11 fusion via alpha channeling with a hybrid fast and thermal proton scheme.

The proton-boron-11 (p-B11) fusion reaction is much harder to harness for commercial power than the easiest fusion reaction, namely, the deuterium and tritium (DT) reaction. The p-B11 reaction requires much higher temperatures, and, even at those higher temperatures, the cross sectionis much smaller. However, as opposed to tritium, the reactants are both abundant and nonradioactive. It is also an aneutronic reaction, thus avoiding radioactivity-inducing neutrons. Economical fusion can only result, however, if the plasma is nearly ignited; in other words if the fusion power is at least nearly equal to the power lost due to radiation and thermal conduction. Because the required temperatures are so high, ignition is thought barely possible for p-B11, with fusion power exceeding the bremsstrahlung power by only around 3%. We show that there is a high upside to changing the natural flow of power in the reactor, putting more power into protons, and less into the electrons. This redirection can be done using waves, which tap the alpha particle power and redirect it into protons through alpha channeling. Using a simple power balance model, we show that such channeling could reduce the required energy confinement time for ignition by a factor of 2.6 when energy is channeled into thermal protons, and a factor of 6.9 when channeled into fast protons near the peak of the reactivity. Thus, alpha channeling could dramatically improve the feasibility of economical p-B11 fusion energy.

Lossless embedding: A visually meaningful image encryption algorithm based on hyperchaos and compressive sensing

A novel visually meaningful image encryption algorithm is proposed based on a hyperchaotic system and compressive sensing (CS), which aims to improve the visual security of steganographic image and decrypted quality. First, a dynamic spiral block scrambling is designed to encrypt the sparse matrix generated by performing discrete wavelet transform (DWT) on the plain image. Then, the encrypted image is compressed and quantified to obtain the noise-like cipher image. Then the cipher image is embedded into the alpha channel of the carrier image in portable network graphics (PNG) format to generate the visually meaningful steganographic image. In our scheme, the hyperchaotic Lorenz system controlled by the hash value of plain image is utilized to construct the scrambling matrix, the measurement matrix and the embedding matrix to achieve higher security. In addition, compared with other existing encryption algorithms, the proposed PNG-based embedding method can blindly extract the cipher image, thus effectively reducing the transmission cost and storage space. Finally, the experimental results indicate that the proposed encryption algorithm has very high visual security.

LRO‐LAMP Survey of Lunar South Pole Cold Traps: Implication for the Presence of Condensed H2O

AbstractWe employ nighttime observations from the Lyman Alpha Mapping Project (LAMP) ultraviolet (UV) spectrograph onboard the Lunar Reconnaissance Orbiter to investigate the presence of condensed water ice within lunar south pole cold traps. This study incorporates LAMP observations between 2009 and 2016, which more than doubles the number of observations from previous LAMP Permanently Shadowed Regions (PSRs) studies. In this study, we focus on cold traps within five south polar craters: Faustini, Shoemaker, Haworth, Cabeus, and Amundsen. An unnamed cold trap region adjacent to Haworth and Shoemaker was also investigated. We present On‐band (148–162 nm), Edge‐band (164–173 nm), and Off‐band (175–190 nm) albedo maps at 4‐km spatial resolution and far‐UV albedo spectra of cold traps at 2‐nm spectral resolution. We find Off‐band to On‐band albedo ratios (the UV wavelength ranges with high and low H2O ice reflectivity, respectively) within cold traps that are consistent with previous PSR studies. Spectral intimate mixture models considering 5 μm H2O grains and a range of regolith grain sizes (20–74 μm) with porosities between 0.3 and 0.75 yields abundances of ∼0.9–4.9 wt.%, where increasing the range of H2O and regolith grain sizes may increase the range of water abundance. We further find generally lower albedos within the cold trap regions than at the surrounding non‐cold trap regions across all wavelengths, consistent with increased regolith porosity within cold traps as suggested by Gladstone et al. (2012), https://doi.org/10.1029/2011JE003913.

A Novel Approach of Virtual Visualization of Cloth Fitting

The concept of a virtual fitting room in real time has been proposed. The interest in online shopping has grown exponentially. When it comes to buying products like shirts that always require knowledge of how the clothes will fit when worn. This is the main reason why so few clothes are being bought online. As a result, a virtual dressing room that tells consumers how clothes fit personally will be a huge asset to online sellers that can give consumers a wide choice. For online marketers, this would be a great way to grow your market. The proposed system consists of a number of tasks, including locating the user's shirt and specifying the color of the user's shirt. In our proposed system we have used Alpha Channel Masking to mask the user's shirt, and libraries such as Numpy, OpenCV. Finally, we conclude that the concept of virtual fitting rooms that we have implemented helps humans in many ways, such as making their purchases easier, saving them time

3D matting: A benchmark study on soft segmentation method for pulmonary nodules applied in computed tomography

Usually, lesions are not isolated but are associated with the surrounding tissues. For example, the growth of a tumour can depend on or infiltrate into the surrounding tissues. Due to the pathological nature of the lesions, it is challenging to distinguish their boundaries in medical imaging. However, these uncertain regions may contain diagnostic information. Therefore, the simple binarization of lesions by traditional binary segmentation can result in the loss of diagnostic information. In this work, we introduce the image matting into the 3D scenes and use the alpha matte, i.e., a soft mask, to describe lesions in a 3D medical image. The traditional soft mask acted as a training trick to compensate for the easily mislabelled or under-labelled ambiguous regions. In contrast, 3D matting uses soft segmentation to characterize the uncertain regions more finely, which means that it retains more structural information for subsequent diagnosis and treatment. The current study of image matting methods in 3D is limited. To address this issue, we conduct a comprehensive study of 3D matting, including both traditional and deep-learning-based methods. We adapt four state-of-the-art 2D image matting algorithms to 3D scenes and further customize the methods for CT images to calibrate the alpha matte with the radiodensity. Moreover, we propose the first end-to-end deep 3D matting network and implement a solid 3D medical image matting benchmark. Its efficient counterparts are also proposed to achieve a good performance-computation balance. Furthermore, there is no high-quality annotated dataset related to 3D matting, slowing down the development of data-driven deep-learning-based methods. To address this issue, we construct the first 3D medical matting dataset. The validity of the dataset was verified through clinicians’ assessments and downstream experiments. The dataset and codes will be released to encourage further research.11Url for codes and dataset: https://github.com/wangsssky/3DMatting.

Graph signal processing and tunicate swarm optimization based image steganography using hybrid chaotic map based image scrambling

This paper put forwards a secure image steganography technique for transmitting sensitive data using a Tunicate swarm optimization algorithm. A hybrid chaotic map-based image scrambling method is proposed with the combination of Tent and Tinkerbell chaotic maps. The scrambling of the secret image provides an added level of security in data hiding. The graph signal processing of cover image using the Meyer wavelet filter is performed to conceal the secret data within the transformed cover image using the alpha blending operation. To discover the optimal value of alpha for alpha blending, the tunicate swarm optimization algorithm is employed. To take the hidden information out at receiver end, the reverse approach is applied. The experimental results revealed that the embedding technique using the tunicate algorithm with a hybrid chaotic map provides better imperceptibility and robustness in comparison to the existing graph signal-based image steganography technique. The investigation is done on PSNR, MSE and SSIM performance parameters for evaluating the proposed technique.

Algorithm of Encrypting Digital Image Using Chaos Neural Network

With the rapid development of digital communication technology, data and image encryption methods will get more and more in-depth research in these aspects. Although the original cipher method can keep the image confidential, because of the storage method of the image data, for the image with too high pixel value, the process of keeping the two-dimensional information confidential and decoding will bring a large amount of computation, and the accuracy of encryption and decryption cannot be guaranteed. This paper presents a visual image encryption algorithm based on the compressed sensing method of the Hopfield chaotic neural network. Different from the existing visual image encryption methods, our method can directly embed the noise like information after encryption into the alpha channel of the carrier image, thus increasing the visual stability of the computer. Finally, through the comparison and analysis of simulation experiments, we find that the image encryption algorithm proposed in this paper has the advantages of wide password space, low risk, good visual stability, good decoding effect, and good robustness.

Coarse Semantic Guided Alpha Matting via Simultaneous Foreground and Background Estimation

To address the "color spill" problem in image matting, an algorithm of simultaneously predicting the foreground, background and alpha matte is proposed, and an automatic matting network combined with coarse semantics is designed. The algorithm method includes two stages: coarse semantics generation and matting via simultaneous foreground and background prediction. In the first stage, a semantic segmentation network is used to obtain intermediate results, and then a coarse semantic information fusion module is applied to roughly estimate the semantic information on multiple scales. In the second stage, an encoder-decoder is used to refine the semantic information and obtain the prediction of foreground, background and alpha matte. Foreground objects can be extracted more accurately through this two-stage network. Also, the obtained alpha matte can be directly used for downstream tasks such as image and video processing. The experimental results on Adobe dataset and the Distinction-646 dataset show that the sum of absolution difference is 42.5 and 50.3, the gradient error is 27.1 and 28.0, respectively. And the details are also more accurate.

EEG Signal Classification of Motor Imagery Right and Left Hand using Common Spatial Pattern and Multilayer Perceptron Back Propagation

The number of people with disabilities is increasing, so it requires bionic devices to replace human motor functions. Brain-Computer Interface (BCI) can be a tool for the bionic device to communicate with the brain. Signal brain or Electroencephalogram (EEG) signal need to classify to drive the corresponding bionic device. This research goal is to classify the imagination of the right and left-hand movements based on the EEG signal. The system design in this research consists of EEG channel selection using Finite Impulse Response (FIR) filter, feature extraction using Common Spatial Pattern (CSP), and classification using Multilayer Perceptron Back Propagation (MLP-BP). The data used a secondary dataset from BCI Competition IV (2b) with 9 research subjects. The research scenario is carried out by trying to use several variations in the number of hidden layer nodes on each EEG channel. Based on the test, the best accuracy for MLP-BP is 68.7% using 24 nodes in the alpha channel.

A Novel Approach of Virtual Visualization of Cloth Fitting

The concept of a virtual fitting room in real time has been proposed. The interest in online shopping has grown exponentially. When it comes to buying products like shirts that always require knowledge of how the clothes will fit when worn. This is the main reason why so few clothes are being bought online. As a result, a virtual dressing room that tells consumers how clothes fit personally will be a huge asset to online sellers that can give consumers a wide choice. For online marketers, this would be a great way to grow your market. The proposed system consists of a number of tasks, including locating the user's shirt and specifying the color of the user's shirt. In our proposed system we have used Alpha Channel Masking to mask the user's shirt, and libraries such as Numpy, OpenCV. Finally, we conclude that the concept of virtual fitting rooms that we have implemented helps humans in many ways, such as making their purchases easier, saving them time

An efficient double-image encryption and hiding algorithm using a newly designed chaotic system and parallel compressive sensing

In this paper, an efficient double-image visually meaningful encryption algorithm combining a new chaotic system, parallel compressive sensing (PCS) and fast Fourier transform (FFT), is proposed. First, Arnold scrambling is performed on two plain images after wavelet transform. Then, the scrambled secret images are compressed by applying a key-controlled partial Hadamard matrix. Finally, the final cipher image with the same resolution as the plain image is acquired by embedding the two encrypted and measured secret images into a carrier image through FFT. To achieve a higher level of security, a new chaotic system combining infinite collapses and a tent map (ICTM) is introduced for disturbing the strong correlation of images and randomly embedding encrypted data into carrier images. The parameters of ICTM are generated by the hash value of the plain images, which enhances the ability of the cryptographic system to resist plaintext attacks. Moreover, the parameters related to the plaintext are embedded into the alpha channel of the visually meaningful cipher image to reduce unnecessary key transmission. The simulation results and performance analysis show that the cryptosystem is feasible and efficient under the condition of guaranteeing all the evaluation indices.

Control quality measurements of alpha and beta counter with low back-ground ALBA 2000 v.2.5.6.

The article provides a systematic approach to quality control of measurements of total alpha-and beta-activity using a counter with a low background ALPHA / BETA COUNTING SYSTEM ALBA (mod. ALBA / LLAB) and software ALBA 2000 v.2.5.6. The purpose is to determine the compliance of these objects with the requirements of regulatory documentation. The spectrometry method is based on the physical concentration of radionuclides from the sample volume, measuring the rate of alpha, and beta radiation of the obtained dry residue of the sample, comparing the sample count rate with calibration values of activity, and calculating the total alpha, beta activity of the sample. The primary means of testing is a counter with a low background ALPHA / BETA COUNTING SYSTEM ALBA 200, the lower limit of measurements of the alpha activity, which is 0.02 Bq/l, and beta activity of 0.1 Bq/l, the relative random uncertainty of the measurement result is 60 % with a confidence level P = 0.95. The efficiency of registration on the alpha channel of 43 %, background on the alpha channel of 0,11 imp./min, for a measurement time of 60,000 s, the efficiency of registration on the beta channel of 30 %, background on the beta channel of 1,9 imp./min, for measurement time 60000 s. As a comparison sample for calculating total alpha activity, a sample with alpha radiation, 241Am (geometry 2π), is used. To calculate the total beta activity, a sample of beta-emitter of potassium sulfate with radionuclide is used at 40K (geometry 2π). The weight of the counting sample ranges from 200 to 1000 mg. The activity values are calculated automatically, using the software ALBA 2000, v.2.5.6. Measurement of the total activity of radionuclides in counting samples using the method should be performed only in calibrated geometries. At the same time, the safety requirements, personnel qualifications, and test conditions must be met. Quality control of measurements in the alpha, beta counter ALBA-2000 v.2.5.6. It is carried out in qualitative and quantitative ways.

Unsupervised Deep Background Matting Using Deep Matte Prior

Background matting is a recently developed image matting approach, with applications to image and video editing. It refers to estimating both the alpha matte and foreground from a pair of images with and without foreground objects. Recent work has applied deep learning to background matting, with very promising performance achieved. However, existing deep models are supervised which require a large dataset with ground truth alpha mattes for training. To avoid the cost of data collection and possible bias in training data, this paper proposes a dataset-free unsupervised deep learning-based approach for background matting. Observing that the local smoothness of alpha matte can be well characterized by the untrained network prior called deep matte prior, we model the foreground and alpha matte using the priors encoded by two generative convolutional neural networks. To avoid possible overfitting during unsupervised learning, a two-stage learning scheme is developed which contains projection-based training and Bayesian post refinement. An alpha-matte-driven initialization scheme is also developed for performance boost. Even without calling external training data, the proposed approach provides competitive performance to recent supervised learning-based methods in the experiments.

Deep collaborative network with alpha matte for precise knee tissue segmentation from MRI

Background and objectivePrecise segmentation of knee tissues from magnetic resonance imaging (MRI) is critical in quantitative imaging and diagnosis. Convolutional neural networks (CNNs), being state of the art, often challenged by the lack of image-specific adaptation, such as low tissue contrasts and structural inhomogeneities, thereby leading to incomplete segmentation results. MethodsThis paper presents a deep learning–based automatic segmentation framework for precise knee tissue segmentation. A novel deep collaborative method is proposed, which consists of an encoder–decoder-based segmentation network in combination with a low rank tensor-reconstructed segmentation network. Low rank reconstruction in MRI tensor sub-blocks is introduced to exploit the morphological variations in knee tissues. To model the tissue boundary regions and effectively utilize the superimposed regions, trimap generation is proposed for defining high, medium and low confidence regions from the multipath CNNs. The secondary path with low rank reconstructed input mitigates the conditions in which the primary segmentation network can potentially fail and overlook the boundary regions. The outcome of the segmentation is solved as an alpha matting problem by blending the trimap with the source input. ResultsExperiments on Osteoarthritis Initiative (OAI) datasets with all the 6 musculoskeletal tissues provide an overall segmentation dice score of 0.8925, where Femoral and Tibial part of cartilage achieving an average dice exceeding 0.9. The volumetric metrics also indicate the superior performances in all tissue compartments. ConclusionsExperiments on Osteoarthritis Initiative (OAI) datasets and a self-prepared scan validate the effectiveness of the proposed method. Inclusion of extra prediction scale allowed the model to distinguish and segment the tissue boundary accurately. We specifically demonstrate the application of the proposed method in a cartilage segmentation–based thickness map for diagnosis purposes.

Effects of short-term food deprivation on catecholamine and metabolic-sensory biomarker gene expression in hindbrain A2 noradrenergic neurons projecting to the forebrain rostral preoptic area: Impact of negative versus positive estradiol feedback.

Hindbrain A2 noradrenergic neurons assimilate estrogenic and metabolic cues. In female mammals, negative- versus positive-feedback patterns of estradiol (E) secretion impose divergent regulation of the gonadotropin-releasing hormone (GnRH)-pituitary-gonadal (HPG) neuroendocrine axis. Current research used retrograde tracing, dual-label immunocytochemistry, single-cell laser-microdissection, and multiplex qPCR methods to address the premise that E feedback modes uniquely affect metabolic regulation of A2 neurons involved in HPG control. Ovariectomized female rats were given E replacement to replicate plasma hormone levels characteristic of positive (high-E dose) or negative (low-E dose) feedback. Animals were either full-fed (FF) or subjected to short-term, e.g., 18-h food deprivation (FD). After FF or FD, rostral preoptic area (rPO)-projecting A2 neurons were characterized by the presence or absence of nuclear glucokinase regulatory protein (nGKRP) immunostaining. FD augmented or suppressed mRNAs encoding the catecholamine enzyme dopamine-beta-hydroxylase (DβH) and the metabolic-sensory biomarker glucokinase (GCK), relative to FF controls, in nGKRP-immunoreactive (ir)-positive A2 neurons from low-E or high-E animals, respectively. Yet, these transcript profiles were unaffected by FD in nGKRP-ir-negative A2 neurons at either E dosage level. FD altered estrogen receptor (ER)-alpha and ATP-sensitive potassium channel subunit sulfonylurea receptor-1 gene expression in nGKRP-ir-positive neurons from low-E, but not high-E animals. Results provide novel evidence that distinct hindbrain A2 neuron populations exhibit altered versus unaffected transmission to the rPO during FD-associated metabolic imbalance, and that the direction of change in this noradrenergic input is controlled by E feedback mode. These A2 cell types are correspondingly distinguished by FD-sensitive or -insensitive GCK, which correlates with the presence versus absence of nGKRP-ir. Further studies are needed to determine how E signal volume regulates neurotransmitter and metabolic sensor responses to FD in GKRP-expressing A2 neurons.

Momentum conservation in current drive and alpha-channeling-mediated rotation drive

Alpha channeling uses waves to extract hot ash from a fusion plasma, transferring energy from the ash to the wave. It has been proposed that this process could create a radial electric field, efficiently driving E×B rotation. However, existing theories ignore the nonresonant particles, which play a critical role in enforcing momentum conservation in quasilinear theory. Because cross field charge transport and momentum conservation are fundamentally linked, this non-consistency throws the rotation drive into question. This paper has two main goals. First, we provide a pedantic and cohesive introduction to the recently developed simple, general, self-consistent quasilinear theory for electrostatic waves that explains the torques which allow for current drive parallel to the magnetic field, and charge extraction across it; a theory that has largely resolved the question of rotation drive by alpha channeling. We show how the theory reveals a fundamental difference between the reaction of nonresonant particles to plane waves that grow in time vs steady-state waves that have a nonuniform spatial structure, allowing rotation drive in the latter case while precluding it in the former, and we review the local and global conservation laws that lead to this result. Second, we provide two new results in support of the theory. First, we provide a novel two-particle Hamiltonian model that rigorously establishes the relationship between charge transport and momentum conservation. Second, we compare the new quasilinear theory to the oscillation-center theories of ponderomotive forces, showing how the latter often obscure the time-dependent nonresonant recoil, but ultimately lead to similar results.

Some Properties of Fuzzy Neutrosophic Generalized Semi Continuous Mapping and Alpha Generalized Continuous Mapping

In the current study, the definition of mapping of fuzzy neutrosophic generalized semi-continuous and fuzzy neutrosophic alpha has generalized mapping as continuous. The study confirmed some theorems regarding such a concept. In the following, it has been found relationships among fuzzy neutrosophic alpha generalized mapping as continuous, fuzzy neutrosophic mapping as continuous, fuzzy neutrosophic alpha mapping as continuous, fuzzy neutrosophic generalized semi mapping as continuous, fuzzy neutrosophic pre mapping as continuous and fuzzy neutrosophic γ mapping as continuous.

Local complexity difference matting based on weight map and alpha mattes

Local complexity difference matting based on weight map and alpha mattes

Estimating Alpha, Beta, and Gamma Diversity Through Deep Learning.

The reliable mapping of species richness is a crucial step for the identification of areas of high conservation priority, alongside other value and threat considerations. This is commonly done by overlapping range maps of individual species, which requires dense availability of occurrence data or relies on assumptions about the presence of species in unsampled areas deemed suitable by environmental niche models. Here, we present a deep learning approach that directly estimates species richness, skipping the step of estimating individual species ranges. We train a neural network model based on species lists from inventory plots, which provide ground truth data for supervised machine learning. The model learns to predict species richness based on spatially associated variables, including climatic and geographic predictors, as well as counts of available species records from online databases. We assess the empirical utility of our approach by producing independently verifiable maps of alpha, beta, and gamma plant diversity at high spatial resolutions for Australia, a continent with highly heterogeneous diversity patterns. Our deep learning framework provides a powerful and flexible new approach for estimating biodiversity patterns, constituting a step forward toward automated biodiversity assessments.