Deep Learning-based Biological Anatomical Landmark Detection in Colonoscopy Videos

Colonoscopy is a standard imaging tool for visualizing the entire gastrointestinal (GI) tract of patients to capture lesion areas. However, it takes the clinicians excessive time to review a large number of images extracted from colonoscopy videos. Thus, automatic detection of biological anatomical landmarks within the colon is highly demanded, which can help reduce the burden of clinicians by providing guidance information for the locations of lesion areas. In this article, we propose a novel deep learning-based approach to detect biological anatomical landmarks in colonoscopy videos. First, raw colonoscopy video sequences are pre-processed to reject interference frames. Second, a ResNet-101 based network is used to detect three biological anatomical landmarks separately to obtain the intermediate detection results. Third, to achieve more reliable localization of the landmark periods within the whole video period, we propose to post-process the intermediate detection results by identifying the incorrectly predicted frames based on their temporal distribution and reassigning them back to the correct class. Finally, the average detection accuracy reaches 99.75%. Meanwhile, the average IoU of 0.91 shows a high degree of similarity between our predicted landmark periods and ground truth. The experimental results demonstrate that our proposed model is capable of accurately detecting and localizing biological anatomical landmarks from colonoscopy videos.

Capturing Event Argument Interaction via A Bi-Directional Entity-Level Recurrent Decoder

Capturing interactions among event arguments is an essential step towards robust event argument extraction (EAE). However, existing efforts in this direction suffer from two limitations: 1) The argument role type information of contextual entities is mainly utilized as training signals, ignoring the potential merits of directly adopting it as semantically rich input features; 2) The argument-level sequential semantics, which implies the overall distribution pattern of argument roles over an event mention, is not well characterized. To tackle the above two bottlenecks, we formalize EAE as a Seq2Seq-like learning problem for the first time, where a sentence with a specific event trigger is mapped to a sequence of event argument roles. A neural architecture with a novel Bi-directional Entity-level Recurrent Decoder (BERD) is proposed to generate argument roles by incorporating contextual entities argument role predictions, like a word-by-word text generation process, thereby distinguishing implicit argument distribution patterns within an event more accurately.

A Trust-Centric Privacy-Preserving Blockchain for Dynamic Spectrum Management in IoT Networks

In this paper, we propose a trust-centric privacy-preserving blockchain for dynamic spectrum access in IoT networks. To be specific, we propose a trust evaluation mechanism to evaluate the trustworthiness of sensing nodes and design a Proof-of-Trust (PoT) consensus mechanism to build a scalable blockchain with high transaction-per-second (TPS). Moreover, a privacy protection scheme is proposed to protect sensors real-time geolocatioin information when they upload sensing data to the blockchain. Two smart contracts are designed to make the whole procedure (spectrum sensing, spectrum auction, and spectrum allocation) run automatically. Simulation results demonstrate the expected computation cost of the PoT consensus algorithm for reliable sensing nodes is low, and the cooperative sensing performance is improved with the help of trust value evaluation mechanism.In addition, incentivization and security are also analyzed, which show that our design not only can encourage nodes participation, but also resist to many kinds of attacks which are frequently encountered in trust-based blockchain systems.

No impact of core-scale magnetic field, turbulence, or velocity gradient on sizes of protostellar disks in Orion A

We compared the sizes and fluxes of a sample of protostellar disks in Orion A measured with the ALMA 0.87 mm continuum data from the VANDAM survey with the physical properties of their ambient environments on the core scale of 0.6 pc estimated with the GBT GAS NH3 and JCMT SCUPOL polarimetric data. We did not find any significant dependence of the disk radii and continuum fluxes on a single parameter on the core scale, such as the non-thermal line width, magnetic field orientation and strength, or magnitude and orientation of the velocity gradient. Among these parameters, we only found a positive correlation between the magnitude of the velocity gradient and the non-thermal line width. Thus, the observed velocity gradients are more likely related to turbulent motion but not large-scale rotation. Our results of no clear dependence of the disk radii on these parameters are more consistent with the expectation from non-ideal MHD simulations of disk formation in collapsing cores, where the disk size is self-regulated by magnetic braking and diffusion, compared to other simulations which only include turbulence and/or a magnetic field misaligned with the rotational axis. Therefore, our results could hint that the non-ideal MHD effects play a more important role in the disk formation. Nevertheless, we cannot exclude the influences on the observed disk size distribution by dynamical interaction in a stellar cluster or amounts of angular momentum on the core scale, which cannot be probed with the current data.

Exploiting Method Names to Improve Code Summarization: A Deliberation Multi-Task Learning Approach

Code summaries are brief natural language descriptions of source code pieces. The main purpose of code summarization is to assist developers in understanding code and to reduce documentation workload. In this paper, we design a novel multi-task learning (MTL) approach for code summarization through mining the relationship between method code summaries and method names. More specifically, since a methods name can be considered as a shorter version of its code summary, we first introduce the tasks of generation and informativeness prediction of method names as two auxiliary training objectives for code summarization. A novel two-pass deliberation mechanism is then incorporated into our MTL architecture to generate more consistent intermediate states fed into a summary decoder, especially when informative method names do not exist. To evaluate our deliberation MTL approach, we carried out a large-scale experiment on two existing datasets for Java and Python. The experiment results show that our technique can be easily applied to many state-of-the-art neural models for code summarization and improve their performance. Meanwhile, our approach shows significant superiority when generating summaries for methods with non-informative names.

Dynamical stellar masses of pre-main sequence stars in Lupus and Taurus obtained with ALMA surveys in comparison with stellar evolutionary models

We analysed archival molecular line data of pre-main sequence (PMS) stars in the Lupus and Taurus star-forming regions obtained with ALMA surveys with an integration time of a few minutes per source. We stacked the data of $^{13}$CO and C$^{18}$O (J = 2-1 & 3-2) and CN (N = 3-2, J = 7/2-5/2) lines to enhance the signal-to-noise ratios, and measured the stellar masses of 45 out of 67 PMS stars from the Keplerian rotation in their circumstellar disks. The measured dynamical stellar masses were compared to the stellar masses estimated from the spectroscopic measurements with seven different stellar evolutionary models. We found that the magnetic model of Feiden (2016) provides the best estimate of the stellar masses in the mass range of $0.6~M_{odot}leq M_{star} leq 1.3~M_{odot}$ with a deviation of $<$0.7$sigma$ from the dynamical masses, while all the other models underestimate the stellar masses in this mass range by 20% to 40%. In the mass range of $<0.6~M_{odot}$, the stellar masses estimated with the magnetic model of Feiden (2016) have a larger deviation ($>2sigma$) from the dynamical masses, and other, non-magnetic stellar evolutionary models of Siess et al. (2000), Baraffe et al. (2015) and Feiden (2016) show better agreements with the dynamical masses with the deviations of 1.4$sigma$ to 1.6$sigma$. Our results show the mass dependence of the accuracy of these stellar evolutionary models.

Improving phase estimation using the number-conserving operations

We propose a theoretical scheme to improve the resolution and precision of phase measurement with parity detection in the Mach-Zehnder interferometer by using a nonclassical input state which is generated by applying a number-conserving generalized superposition of products (GSP) operation, (saa^{{dag}}+ta^{{dag}}a)^{m} with s^2+t^2=1, on two-mode squeezed vacuum (TMSV) state. The nonclassical properties of the proposed GSP-TMSV are investigated via average photon number (APN), anti-bunching effect, and degrees of two-mode squeezing. Particularly, our results show that both higher-order m GSP operation and smaller parameter s can increase the total APN, which leads to the improvement of quantum Fisher information. In addition, we also compare the phase measurement precision with and without photon losses between our scheme and the previous photon subtraction/addition schemes. It is found that our scheme, especially for the case of s=0, has the best performance via the enhanced phase resolution and sensitivity when comparing to those previous schemes even in the presence of photon losses. Interestingly, without losses, the standard quantum-noise limit (SQL) can always be surpassed in our our scheme and the Heisenberg limit (HL) can be even achieved when s=0.5,1 with small total APNs. However, in the presence of photon losses, the HL cannot be beaten, but the SQL can still be overcome particularly in the large total APN regimes. Our results here can find important applications in quantum metrology.

The JCMT BISTRO survey: alignment between outflows and magnetic fields in dense cores/clumps

We compare the directions of molecular outflows of 62 low-mass Class 0 and I protostars in nearby (<450 pc) star-forming regions with the mean orientations of the magnetic fields on 0.05-0.5 pc scales in the dense cores/clumps where they are embedded. The magnetic field orientations were measured using the JCMT POL-2 data taken by the BISTRO-1 survey and from the archive. The outflow directions were observed with interferometers in the literature. The observed distribution of the angles between the outflows and the magnetic fields peaks between 15 and 35 degrees. After considering projection effects, our results could suggest that the outflows tend to be misaligned with the magnetic fields by 50+/-15 degrees in three-dimensional space and are less likely (but not ruled out) randomly oriented with respect to the magnetic fields. There is no correlation between the misalignment and the bolometric temperatures in our sample. In several sources, the small-scale (1000-3000 au) magnetic fields is more misaligned with the outflows than their large-scale magnetic fields, suggesting that the small-scale magnetic field has been twisted by the dynamics. In comparison with turbulent MHD simulations of core formation, our observational results are more consistent with models in which the energy densities in the magnetic field and the turbulence of the gas are comparable. Our results also suggest that the misalignment alone cannot sufficiently reduce the efficiency of magnetic braking to enable formation of the observed number of large Keplerian disks with sizes larger than 30-50 au.

Cross-Modal Weighting Network for RGB-D Salient Object Detection

Depth maps contain geometric clues for assisting Salient Object Detection (SOD). In this paper, we propose a novel Cross-Modal Weighting (CMW) strategy to encourage comprehensive interactions between RGB and depth channels for RGB-D SOD. Specifically, three RGB-depth interaction modules, named CMW-L, CMW-M and CMW-H, are developed to deal with respectively low-, middle- and high-level cross-modal information fusion. These modules use Depth-to-RGB Weighing (DW) and RGB-to-RGB Weighting (RW) to allow rich cross-modal and cross-scale interactions among feature layers generated by different network blocks. To effectively train the proposed Cross-Modal Weighting Network (CMWNet), we design a composite loss function that summarizes the errors between intermediate predictions and ground truth over different scales. With all these novel components working together, CMWNet effectively fuses information from RGB and depth channels, and meanwhile explores object localization and details across scales. Thorough evaluations demonstrate CMWNet consistently outperforms 15 state-of-the-art RGB-D SOD methods on seven popular benchmarks.

Discriminator Contrastive Divergence: Semi-Amortized Generative Modeling by Exploring Energy of the Discriminator

Generative Adversarial Networks (GANs) have shown great promise in modeling high dimensional data. The learning objective of GANs usually minimizes some measure discrepancy, textit{e.g.}, $f$-divergence~($f$-GANs) or Integral Probability Metric~(Wasserstein GANs). With $f$-divergence as the objective function, the discriminator essentially estimates the density ratio, and the estimated ratio proves useful in further improving the sample quality of the generator. However, how to leverage the information contained in the discriminator of Wasserstein GANs (WGAN) is less explored. In this paper, we introduce the Discriminator Contrastive Divergence, which is well motivated by the property of WGANs discriminator and the relationship between WGAN and energy-based model. Compared to standard GANs, where the generator is directly utilized to obtain new samples, our method proposes a semi-amortized generation procedure where the samples are produced with the generators output as an initial state. Then several steps of Langevin dynamics are conducted using the gradient of the discriminator. We demonstrate the benefits of significant improved generation on both synthetic data and several real-world image generation benchmarks.