Search

Neural computations are currently investigated using two separate approaches: sorting neurons into functional sub-populations, or examining the low-dimensional dynamics of collective activity. Whether and how these two aspects interact to shape computations is currently unclear. Using a novel approach to extract computational mechanisms from networks trained on neuroscience tasks, here we show that the dimensionality of the dynamics and sub-population structure play fundamentally complementary roles. While various tasks can be implemented by increasing the dimensionality in networks with fully random population structure, flexible input-output mappings instead require a non-random population structure that can be described in terms of multiple sub-populations. Our analyses revealed that such a sub-population structure enables flexible computations through a mechanism based on gain-controlled modulations that flexibly shape the collective dynamics. Our results lead to task-specific predictions for the structure of neural selectivity, inactivation experiments, and for the implication of different neurons in multi-tasking.

Destruction and death of neurons can lead to neurodegenerative diseases. One possible way to treat neurodegenerative diseases and damage of the nervous system is replacing damaged and dead neurons by cell transplantation. If new neurons can replace the lost neurons, patients may be able to regain the lost functions of memory, motor, and so on. Therefore, acquiring neurons conveniently and efficiently is vital to treat neurological diseases. In recent years, studies on reprogramming human fibroblasts into neurons have emerged one after another, and this paper summarizes all these studies. Scientists find small molecules and transcription factors playing a crucial role in reprogramming and inducing neuron production. At the same time, both the physiological microenvironment in vivo and the physical and chemical factors in vitro play an essential role in the induction of neurons. Therefore, this paper summarized and analyzed these relevant factors. In addition, due to the unique advantages of physical factors in the process of reprogramming human fibroblasts into neurons, such as safe and minimally invasive, it has a more promising application prospect. Therefore, this paper also summarizes some successful physical mechanisms of utilizing fibroblasts to acquire neurons, which will provide new ideas for somatic cell reprogramming.

The vortex electromagnetic (EM) wave carrying orbital angular momentum (OAM) offers a new degree of freedom for synthetic aperture radar (SAR) imaging. Because vortex EM waves have helical wavefronts, the vortex EM wave-based SAR echo contains the three-dimensional (3D) information of the target. In this paper, an OAM-based synthetic aperture radar interferometry (InSAR) technique is proposed to obtain 3D target information accurately without the existence of baseline. First, a vortex EM waves' SAR imaging model is established, and an improved range-Doppler algorithm is proposed correspondingly. Subsequently, the scheme of the OAM-based InSAR without the physical baseline is proposed. Compared with the conventional InSAR, the OAM-based InSAR can avoid the baseline decorrelation and reduce the requirement of the platform. Besides, the processing procedure of the OAM-based InSAR is simplified, which avoids the image registration and the interferogram flattening. The simulation results demonstrate the effectiveness of the proposed technique. Besides, the height estimation accuracy of the OAM-based InSAR was analyzed, in terms of interferometric phase error and OAM mode error. The height estimation accuracy can be improved by increasing the OAM mode difference appropriately. The OAM, which is completely independent of time, frequency, and polarization, offers a new scheme for the InSAR.

Active control of terahertz waves is critical to the development of terahertz devices. Two-dimensional materials with excellent optical properties provide more choices for opto-electrical devices due to the advancement in their preparation technology. We proposed a hybrid structure of a metamaterial/monolayer MoS2/Si and investigated its optical properties in the terahertz range. The plasmon-induced transparency (PIT) effect was observed in the transmission spectra, resulting from the near-field coupling of two bright modes. According to the simulated results, this phenomenon confirmed its dependency on the length of the cutwire and the distance between DSSRs. Furthermore, an external optical field supported by a 1064 nm laser could exert a switch effect on the sample. The resonances of the PIT metamaterial disappeared when the optical power was further increased, as the excited carriers in the MoS2/Si substrate blocked the coupling effect. In addition, the experimental results indicated that the PIT metamaterial enhanced the interaction of infrared light with the monolayer MoS2/Si substrate.

For large-scale vision tasks in biomedical images, the labeled data is often limited to train effective deep models. Active learning is a common solution, where a query suggestion method selects representative unlabeled samples for annotation, and the new labels are used to improve the base model. However, most query suggestion models optimize their learnable parameters only on the limited labeled data and consequently become less effective for the more challenging unlabeled data. To tackle this, we propose a two-stream active query suggestion approach. In addition to the supervised feature extractor, we introduce an unsupervised one optimized on all raw images to capture diverse image features, which can later be improved by fine-tuning on new labels. As a use case, we build an end-to-end active learning framework with our query suggestion method for 3D synapse detection and mitochondria segmentation in connectomics. With the framework, we curate, to our best knowledge, the largest connectomics dataset with dense synapses and mitochondria annotation. On this new dataset, our method outperforms previous state-of-the-art methods by 3.1% for synapse and 3.8% for mitochondria in terms of region-of-interest proposal accuracy. We also apply our method to image classification, where it outperforms previous approaches on CIFAR-10 under the same limited annotation budget. The project page is https://zudi-lin.github.io/projects/#two_stream_active.

An amino-functionalized water-retaining agent (PPM40-NH2) and a water-retaining agent lacking functionalization (PPM40) were prepared to investigate their selective Cd(II) removal properties. The affinity for Cd(II) was increased after introducing the amino functional group. The SFCd/Ca of PPM40-NH2 was 3–7 fold higher than that of PPM40 although the CCa was 100–1000 times higher than that of CCd. The chelating effect of –COOH and –NH2 in PPM40-NH2 corresponded well with the selective adsorption of Cd(II), as confirmed by simulations. A seed germination experiment demonstrated that PPM40-NH2 could act as a filter of heavy metals to supply “clean” water for the roots of the pakchoi plant. The leaching experiment indicated that PPM40-NH2 is an efficient in situ fixation agent for the remediation of agricultural soil contaminated with Cd(II) in the presence of high concentrations of Ca(II) salts.

Tomographic Synthetic Aperture Radar (TomoSAR) has become a competitive remote sensing method of three-dimensional (3-D) reconstruction over urban areas. Raw TomoSAR point clouds are inevitably noise-corrupted, which would severely obstruct the reconstruction of building structures. Whereas, data segmentation and parameter tuning are required in current methods of 3-D building reconstruction, which influences the precision and efficiency of the reconstruction process. In this paper we propose a novel method using neural networks to reconstruct 3-D building structures from TomoSAR data. By using the proposed method, the precise 3-D surface of the building structure can be retrieved quickly. The proposed method also performs commendably in point cloud denoising. More importantly, our method achieves full automation of the reconstruction process, which does not require data segmentation or complex parameter adjusting. Experiment results demonstrate the effectiveness of the proposed method.

Foreign object debris (FOD) surveillance radar is used to detect the potential objects which threats the safety of flights. Since the performance of ground based surveillance radar is inevitably restricted by the clutter even the echoes from the runway and grassland of the airport are much stronger than that from tiny FOD, current FOD detection methods are mainly based on clutter modeling and uses backscattering coefficients. In this paper we propose a novel FOD clutter modeling method, which generates the ground distributed clutter by using the back scattering coefficient spatial correlated model. In this model, the amplitude of the clutter is non-Rayleigh distributed and the power spectrum density obeys the Gaussian distribution. Concretely, the runway clutter amplitude applies Weibull distribution and the grassland clutter amplitude adopts Log-normal distribution. The proposed method builds the model under the full consideration of low grazing angle geometry relations and characteristics of FOD surveillance radar. The experiments on the millimeter wave FOD radar experiment data demonstrate the effectiveness of the model. This proposed modeling method provides a fulfilling acquisition for further targets detection process.

Transmission and dielectric properties of non-illuminated and illuminated barium titanate thin film were investigated by time-domain terahertz spectrometer and tuned by an electric field. Compared with the sample illuminated by 532 nm laser, the electric field could realize a higher modulation in the phase of the transmission for the nonilluminated sample. Results displayed that when the sample was illuminated by green laser, optical field suppressed the effect of electric field on transmission and dielectric properties of BTO thin film, because excited carriers from the substrate and ferroelectric thin film would form a built-in electric field to shield the external electric field. In addition, optical field and electrical field would play different roles in the lattice dynamics of the ferroelectric thin film.

The transmission and dielectric spectra of ferroelectric STO/PT superlattice on Si substrate under simultaneous external optical and electric field were investigated and compared at room temperature. Results found that when with an optical field, the electric field realized an effective modulation on the transmission, which displayed a diode property. In addition, a comprehensive model combined with Debye relaxation and Lorentz model was used to analyze the dielectric spectra, variation of the soft mode with external field was put emphasis on exploring.

The dielectric property of ferroelectric barium titanate (BaTiO3) thin film with optical field was investigated by terahertz time-domain spectroscopy at room temperature. Experimental results showed that dielectric constant of BTO film was increased with the optical pump powers, and tunability of the real part of dielectric constant could be reached to74%. The reason of realizing high modulation depth could be explained as photorefractive and photothermal effects. Furthermore, the variation of refractive index displayed a monotonically increase with the optical powers.

The dielectric spectra of ferroelectric Pb(Zr0.52Ti0.48)O3/SrTiO3/PbTiO3 (PZT/STO/PT) superlattice under an external optical field were investigated at room temperature by time-domain terahertz (THz) spectroscopy. Results showed that the real and imaginary part of permittivity were increased roughly upon application of an external optical field, and tuned by 78.16% and 85.65% respectively. It could be interpreted as softening of the soft mode and increasing of the damping coefficient and oscillator strength. Furthermore, the central mode was observed in the tricolor thin film, and the frequency of the central mode was also increased along with the optical power.

Tomographic SAR (TomoSAR) is a remote sensing technique that extends the conventional two-dimensional (2-D) synthetic aperture radar (SAR) imaging principle to three-dimensional (3-D) imaging. It produces 3-D point clouds with unavoidable noise that seriously deteriorates the quality of 3-D imaging and the reconstruction of buildings over urban areas. However, existing methods for TomoSAR point cloud processing notably rely on data segmentation, which influences the processing efficiency and denoising performance to a large extent. Inspired by regression analysis, in this paper, we propose an automatic method using neural networks to regularize the 3-D building structures from TomoSAR point clouds. By changing the point heights, the surface points of a building are refined. The method has commendable performance on smoothening the building surface, and keeps a precise preservation of the building structure. Due to the regression mechanism, the method works in a high automation level, which avoids data segmentation and complex parameter adjustment. The experimental results demonstrate the effectiveness of our method to denoise and regularize TomoSAR point clouds for urban buildings.

An active multifunctional terahertz modulator based on plasmon-induced transparency (PIT) metasurface under the effect of external infrared light was investigated theoretically and experimentally. A distinct transparency window, which resulted from the near-field coupling between two resonators, could be observed in the transmission spectra. Experimental results showed a phenomenon infrared light induced blue shift on the both resonator with increasing optical powers. When the optical power was tuned from 0 mW to 400 mW, the amplitude tunability of transmission at transparency window reached to 34.01%, much larger than that at the two resonance frequencies. Moreover, the phase tunability of the transmission at 0.98 THz reached to 31.35%. Meanwhile, the amplitude variation was limited to 10%. Furthermore, a coupled Lorentz oscillator model was adopted to analyze the near-field interaction of the resonances. Experimental results were in good agreement with the analytical fitting results.

The characteristics of the electromagnetically induced transparency (EIT) analog exposed under different illumination powers have been investigated theoretically and experimentally. The EIT analog is composed of a fixed aluminum structure fabricated on the silicon substrate. It was found that the resonance degree of the transparent window displayed a decreasing trend, and a blueshift phenomenon emerged by increasing the powers of the laser. Similarly, the properties of the time delay under different illumination powers have also been researched. The realization of the tuning effect may provide a possible choice for the modulation of the slow light devices.

We present a multimodal science simulation, including visual and auditory (descriptions, sound effects, and sonifications) display. The design of each modality is described, as well as evaluation with learners with and without visual impairments. We conclude with challenges and opportunities at the intersection of multiple modalities.

Recently, a novel scheme called grayscale-invariance reversible data hiding (RDH) has been developed, in which the generated color marked image will have the same grayscale as that of the host color image. Due to the property of grayscale invariance, for many cases, the further applications and image processing of the marked image will not be interfered with. However, the performance of current grayscale-invariance RDH is unsatisfactory, and the corresponding theoretical model is also lacking. For grayscale-invariance RDH, the modifications of the three color channels are interrelated. By decomposing the joint modification and associating the modification distortions of the green scales and blue scales with those of the red scales, we first approximate the modification-independent rate-distortion model for grayscale-invariance RDH. Subsequently, the optimal transition probability matrix can be estimated, according to which we implement the theoretically proven optimal modification, i.e., recursive code construction (RCC), to finish message embedding. The experimental results show that the presented method extends the applications of RCC and can significantly improve the performance of grayscale-invariance RDH.

Hydroxypropyl guar gum (HPG) is widely used in the hydraulic fracturing of sand rock gas reservoirs. The HPG residue can severely damage the rock surface of the reservoirs owing to the interaction between the HPG molecular chain and the rock surface, which results in adsorption and retention of HPG. To confirm the interaction between HPG and rock, the adsorption properties of HPG on a rock surface were investigated. The dependence of HPG adsorption on HPG concentration, pH, and temperature is discussed, and the adsorption force between HPG and quartz is also studied. With an increase in HPG concentration and pH, the amount of adsorbed HPG increases. The amount of adsorption increases first with temperature and then decreaseds. According to infrared spectroscopy and Zeta potential results, there is a weak chemical bond between HPG and quartz particles. Hydrogen bonding is the main force, which results in the adsorption and retention of HPG in porous media.

Emerging evidence suggests that neuro-inflammation begins early and drives the pathogenesis of Alzheimer's disease (AD), and anti-inflammatory therapies are under clinical development. However, several anti-inflammatory compounds failed to improve memory in clinical trials, indicating that reducing inflammation alone might not be enough. On the other hand, neuro-inflammation is implicated in a number of mental disorders which share the same therapeutic targets. Based on these observations, we screened a batch of genes related with mental disorder and neuro-inflammation in a classical olfactory conditioning in an amyloid beta (Aβ) overexpression fly model. A Smoothened (SMO) mutant was identified as a genetic modifier of Aβ toxicity in 3-min memory and downregulation of SMO rescued Aβ-induced 3-min and 1-h memory deficiency. Also, Aβ activated innate inflammatory response in fly by increasing the expression of antimicrobial peptides, which were alleviated by downregulating SMO. Furthermore, pharmaceutical administration of a SMO antagonist LDE rescued Aβ-induced upregulation of SMO in astrocytes of mouse hippocampus, improved memory in Morris water maze (MWM), and reduced expression of astrocyte secreting pro-inflammatory factors IL-1β, TNFα and the microglia marker IBA-1 in an APP/PS1 transgenic mouse model. Our study suggests that SMO is an important conserved modulator of Aβ toxicity in both fly and mouse models of AD.

This paper presents the gradient-guided image super-resolution reconstruction for terahertz imaging to improve the image quality, taking advantage of super-resolution reconstruction based on adaptive super-pixel gradient field transform. Moreover, spatial entropy-based enhancement and a bilateral filter are introduced to ensure better performance of the reconstruction. Furthermore, we compare the performance of reconstruction operated on terahertz images with frequencies of 0.1 THz, 0.3 THz, 0.5 THz, and 0.7 THz. Experimental results demonstrate that this method successfully improves the image quality and reconstruct high-resolution images from low-resolution images with the peak signal-to-noise ratio and structural similarity index improved. In addition, the signal frequency and intensity are demonstrated to affect the performance of reconstruction.

A hydrophobic polyacrylamide (HPAM) was synthesized by aqueous free-radical chain polymerization to understand the adsorption action of the polymer on the calcite during operating stimulation. The polymer was designed as acrylamide (AM)/2-acrylamide-2-methylpropanesulfonic acid (AMPS)/2-methacryloyloxyethyl 12-alkyl dimethyl ammonium bromide (MADA). The structure of the polymer was characterized by Fourier transform infrared spectroscopy. The adsorption properties of the polyacrylamide onto pure calcite was investigated. When we compared the ζ potentials of the calcite particles in the presence and absence of HPAM, the adsorbed amount was affected by different factors, and the adsorption behaviors were examined to determine the adsorption layer. The results of the ζ potential measurements indicated that the particles were negatively charged between pHs of 6 and 12, and the presence of HPAM did not reverse that. Equilibrium adsorption studies showed that the adsorbed amount was affected by the concentration of HPAM, liquid-to-solid ratio, calcite particle size, concentration of background ions, and temperature, which were related the adsorption behaviors of HPAM. Hydrogen bonds between HPAM and calcite were important because the adsorbed amount was significantly reduced when the hydrogen bonds were broken by urea. It is necessary to develop a more active reagent that can break the hydrogen bonds and improve the effect of hydraulic fracturing. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017, 134, 45314.

Optical properties of ferroelectric superlattice film with different optical powers were characterized by THz spectroscopy. Tunability of transmission could be tuned by 90.1%. Refractive index and absorption coefficient was also varied along with it.

Dielectric spectra of Barium Titanate under an electric field were investigated by time-domain terahertz spectroscopy. The real part of permittivity was increased with the external field while the imaginary part displayed negligible variations. Besides, tunability of dielectric constant reached to 14.8% when the electric field was tuned to 45 V.