Development of an In-Pipe Inspection Robot for Large-Diameter Water Pipes.

This paper describes the development of an in-pipe inspection robot system designed for large-diameter water pipes. The robot is equipped with a Magnetic Flux Leakage (MFL) sensor module. The robot system is intended for pipes with diameters ranging from 900 mm to 1200 mm. The structure of the in-pipe inspection robot consists of the front and rear driving parts, with the inspection module located centrally. The robot is powered by 22 motors, including eight wheels with motors positioned at both the bottom and the top for propulsion. To ensure that the robot's center aligns with that of the pipeline during operation, lifting units have been incorporated. The robot is equipped with cameras and LiDAR sensors at the front and rear to monitor the internal environment of the pipeline. Pipeline inspection is conducted using the MFL inspection modules, and the robot's driving mechanism is designed to execute spiral maneuvers while maintaining contact with the pipeline surface during rotation. The in-pipe inspection robot is configured with wireless communication modules and batteries, allowing for wireless operation. Following its development, the inspection robot underwent driving experiments in actual pipelines to validate its performance. The field test bed used for these experiments is approximately 1 km in length. Results from the driving experiments on the field test bed confirmed the robot's ability to navigate various curvatures and obstacles within the pipeline. It is posited that the use of the developed in-pipe inspection robot can reduce economic costs and enhance the safety of inspectors when examining aging pipes.

Neural Architecture Search Survey: A Computer Vision Perspective.

In recent years, deep learning (DL) has been widely studied using various methods across the globe, especially with respect to training methods and network structures, proving highly effective in a wide range of tasks and applications, including image, speech, and text recognition. One important aspect of this advancement is involved in the effort of designing and upgrading neural architectures, which has been consistently attempted thus far. However, designing such architectures requires the combined knowledge and know-how of experts from each relevant discipline and a series of trial-and-error steps. In this light, automated neural architecture search (NAS) methods are increasingly at the center of attention; this paper aimed at summarizing the basic concepts of NAS while providing an overview of recent studies on the applications of NAS. It is worth noting that most previous survey studies on NAS have been focused on perspectives of hardware or search strategies. To the best knowledge of the present authors, this study is the first to look at NAS from a computer vision perspective. In the present study, computer vision areas were categorized by task, and recent trends found in each study on NAS were analyzed in detail.

Visual question answering based on local-scene-aware referring expression generation.

Visual question answering requires a deep understanding of both images and natural language. However, most methods mainly focus on visual concept; such as the relationships between various objects. The limited use of object categories combined with their relationships or simple question embedding is insufficient for representing complex scenes and explaining decisions. To address this limitation, we propose the use of text expressions generated for images, because such expressions have few structural constraints and can provide richer descriptions of images. The generated expressions can be incorporated with visual features and question embedding to obtain the question-relevant answer. A joint-embedding multi-head attention network is also proposed to model three different information modalities with co-attention. We quantitatively and qualitatively evaluated the proposed method on the VQA v2 dataset and compared it with state-of-the-art methods in terms of answer prediction. The quality of the generated expressions was also evaluated on the RefCOCO, RefCOCO+, and RefCOCOg datasets. Experimental results demonstrate the effectiveness of the proposed method and reveal that it outperformed all of the competing methods in terms of both quantitative and qualitative results.

Cyclocurcumin, an Antivasoconstrictive Constituent of Curcuma longa (Turmeric).

Despite the increasing attention on the therapeutic potential of Curcuma longa (turmeric), the biological activities of curcuminoids other than curcumin are not well understood. Here, we investigated antivasoconstrictive activities of C. longa extract and its ingredients using freshly isolated rat aortic rings. C. longa extract significantly suppressed agonist-stimulated vasoconstriction, and cyclocurcumin was found to be the most potent (IC50 against phenylephrine-induced vasoconstriction: 14.9 ± 1.0 µM) among the 10 tested ingredients including four curcuminoids. Cyclocurcumin significantly inhibited contraction of vascular smooth muscle, which was mediated by the suppression of myosin-light-chain phosphorylation and calcium influx via the L-type calcium channel. The inhibitory effect of cyclocurcumin was observed to be reversible and without cytotoxicity. Taken together, we demonstrated that cyclocurcumin, a bioactive ingredient in C. longa, may have a therapeutic potential as a novel antivasoconstrictive natural product.

Lead optimization of a novel series of imidazo[1,2-a]pyridine amides leading to a clinical candidate (Q203) as a multi- and extensively-drug-resistant anti-tuberculosis agent.

A critical unmet clinical need to combat the global tuberculosis epidemic is the development of potent agents capable of reducing the time of multi-drug-resistant (MDR) and extensively-drug-resistant (XDR) tuberculosis therapy. In this paper, we report on the optimization of imidazo[1,2-a]pyridine amide (IPA) lead compound 1, which led to the design and synthesis of Q203 (50). We found that the amide linker with IPA core is very important for activity against Mycobacterium tuberculosis H37Rv. Linearity and lipophilicity of the amine part in the IPA series play a critical role in improving in vitro and in vivo efficacy and pharmacokinetic profile. The optimized IPAs 49 and 50 showed not only excellent oral bioavailability (80.2% and 90.7%, respectively) with high exposure of the area under curve (AUC) but also displayed significant colony-forming unit (CFU) reduction (1.52 and 3.13 log10 reduction at 10 mg/kg dosing level, respectively) in mouse lung.

Discovery of Q203, a potent clinical candidate for the treatment of tuberculosis.

New therapeutic strategies are needed to combat the tuberculosis pandemic and the spread of multidrug-resistant (MDR) and extensively drug-resistant (XDR) forms of the disease, which remain a serious public health challenge worldwide. The most urgent clinical need is to discover potent agents capable of reducing the duration of MDR and XDR tuberculosis therapy with a success rate comparable to that of current therapies for drug-susceptible tuberculosis. The last decade has seen the discovery of new agent classes for the management of tuberculosis, several of which are currently in clinical trials. However, given the high attrition rate of drug candidates during clinical development and the emergence of drug resistance, the discovery of additional clinical candidates is clearly needed. Here, we report on a promising class of imidazopyridine amide (IPA) compounds that block Mycobacterium tuberculosis growth by targeting the respiratory cytochrome bc1 complex. The optimized IPA compound Q203 inhibited the growth of MDR and XDR M. tuberculosis clinical isolates in culture broth medium in the low nanomolar range and was efficacious in a mouse model of tuberculosis at a dose less than 1 mg per kg body weight, which highlights the potency of this compound. In addition, Q203 displays pharmacokinetic and safety profiles compatible with once-daily dosing. Together, our data indicate that Q203 is a promising new clinical candidate for the treatment of tuberculosis.

Hardware-efficient methods for elimination of signal distortion in sigma-delta-based ultrasound beamforming.

An efficient ultrasound beamforming algorithm based on sigma-delta modulation (SDM) is proposed, which performs dynamic focusing at the Nyquist rate as in conventional delay-sum beamformers. Specifically, a block of consecutive SDM samples for each focal point is selected on each channel according to the focusing delay. The data blocks for all channels are then added up on a bit-to-bit basis and the resulting block of multibit samples is fed to the demodulation filter to produce the focused signal. Since the SDM samples in each data block are not altered at all, the proposed method does not cause any signal distortion. Two hardware structures are also proposed for minimum hardware implementation of the proposed error-free SDM beamforming method. It is verified that compared to the delay-sum beamformer using 10-bit ADCs, the proposed algorithm provides almost the same image quality with a greatly reduced hardware complexity. In addition, it can be implemented with a lower hardware complexity than that of the conventional SDM beamformer when the number of channels is 64 or less.

Real-time 3D imaging methods using 2D phased arrays based on synthetic focusing techniques.

A fast 3D ultrasound imaging technique using a 2D phased array transducer based on the synthetic focusing method for nondestructive testing or medical imaging is proposed. In the proposed method, each column of a 2D array is fired successively to produce transverse fan beams focused at a fixed depth along a given longitudinal direction and the resulting pulse echoes are received at all elements of a 2D array used. After firing all column arrays, a frame of high-resolution image along a given longitudinal direction is obtained with dynamic focusing employed in the longitudinal direction on receive and in the transverse direction on both transmit and receive. The volume rate of the proposed method can be increased much higher than that of the conventional 2D array imaging by employing an efficient sparse array technique. A simple modification to the proposed method can further increase the volume scan rate significantly. The proposed methods are verified through computer simulations.