An Efficient Motion Adjustment Method for a Dual-Arm Transfer Robot Based on a Two-Level Neural Network and a Greedy Algorithm

Although tendon-driven continuum manipulators have been extensively researched, how to realize tip contact force sensing in a more general and efficient way without increasing the diameter is still a challenge. Rather than use a complex modeling approach, this paper proposes a general tip contact force-sensing method based on a recurrent neural network that takes the tendons’ position and tension as the input of a recurrent neural network and the tip contact force of the continuum manipulator as the output and fits this static model by means of machine learning so that it may be used as a real-time contact force estimator. We also designed and built a corresponding three-degree-of-freedom contact force data acquisition platform based on the structure of a continuum manipulator designed in our previous studies. After obtaining training data, we built and compared the performances of a multi-layer perceptron-based contact force estimator as a baseline and three typical recurrent neural network-based contact force estimators through TensorFlow framework to verify the feasibility of this method. We also proposed a manually decoupled sub-estimators algorithm and evaluated the advantages and disadvantages of those two methods.

An investigation on the relationship between grip, push and contact forces applied to a tool handle

Automated image classification is an essential task of the computer vision field. The tagging of images into a set of predefined groups is referred to as image classification. The implementation of computer vision to automate image classification would be beneficial because manual image evaluation and identification can be time-consuming, particularly when there are many images of different classes. Deep learning approaches are proven to overperform existing machine learning techniques in many fields in recent years, and computer vision is one of the most notable examples. The very deep neural network (VDCNN) is a powerful deep learning model for image classification, and this paper examines it briefly using MNIST handwritten digit dataset. This dataset is used to prove the efficacy of very deep neural networks over other deep learning models. The proposed study aims to comprehend the very deep neural network architecture used to accomplish a handwritten digit recognition task. The feasibility of the proposed model is evaluated using mean accuracy, validation accuracy, and standard deviation. The study results of the very deep neural network model are compared to a convolutional neural network and convolutional neural network with batch normalization. According to the results of the comparison study, very deep neural networks achieve high accuracy of 99.1% for a handwritten dataset. The outcome of the proposed work is used to interpret how well a very deep neural network performs when compared to the other two models of deep neural networks. This proposed architecture may be used to automate the classification of handwritten digits dataset.

In general, the contact force, which is used in controller design for robotic compliant motion, can be either explicitly calculated through complicated computation or measured by expensive force sensors. A force control strategy using the estimated contact force is proposed. The estimation of the contact force and states is primarily based on their previous observer design. It is shown that the controller achieves asymptotic stability.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Video hyperlinking represents a classical example of multimodal problems. Common approaches to such problems are early fusion of the initial modalities and crossmodal translation from one modality to the other. Recently, deep neural networks, especially deep autoencoders, have proven promising both for crossmodal translation and for early fusion via multimodal embedding. A particular architecture, bidirectional symmetrical deep neural networks, have been proven to yield improved multimodal embeddings over classical autoencoders, while also being able to perform crossmodal translation. In this work, we focus firstly at evaluating good single-modal continuous representations both for textual and for visual information. Word2Vec and paragraph vectors are evaluated for representing collections of words, such as parts of automatic transcripts and multiple visual concepts, while different deep convolutional neural networks are evaluated for directly embedding visual information, avoiding the creation of visual concepts. Secondly, we evaluate methods for multimodal fusion and crossmodal translation, with different single-modal pairs, in the task of video hyperlinking. Bidirectional (symmetrical) deep neural networks were shown to successfully tackle downsides of multimodal autoencoders and yield a superior multimodal representation. In this work, we extensively tests them in different settings, with different single-modal representations, within the context of video-hyperlinking. Our novel bidirectional symmetrical deep neural networks are compared to classical autoencoders and are shown to yield significantly improved multimodal embeddings that significantly (alpha=0.0001) outperform multimodal embeddings obtained by deep autoencoders with an absolute improvement in precision at 10 of 14.1% when embedding visual concepts and automatic transcripts and an absolute improvement of 4.3% when embedding automatic transcripts with features obtained with very deep convolutional neural networks, yielding 80% of precision at 10.

Contact behaviors of medical devices, such as guidewires and catheters, are critical in endovascular surgeries. In this work, a new method to predict adhesive contact force between catheter and vascular artery is presented. Multi-asperity adhesion on the surface of vascular artery, deformation of asperity and deformation of vascular substrate are all considered. The single asperity behavior is described with Johnson-Kendall-Roberts (JKR) contact model. The multi-asperity behavior is based on Greenwood-Williamson (GW) asperity model. Vascular substrate is considered as elastic bulk substrate and its deformation is determined with Hertzian pressure from asperity on a circular region on the elastic half space. The model shows that the deformation of vascular substrate accounts for the majority of the total contact deformation and significantly affects the predicted contact force. The model is verified with published experimental data. The comparison shows that the model produces very accurate prediction of contact force between catheter and vascular artery when the contact force is compressive. Parametric analysis based on asperity topography is carried out. The analysis shows that the diameter of the circular region of the interface between asperity and vascular substrate has more significant effect on the estimation of contact force than the radius of asperity. Further validation of prediction accuracy of the model under experiment is needed.

Cable-driven continuum robots are widely used for endoluminal intervention because of their dexterity and shape conforming steerability. However, body contact between the continuum robot and its surrounding anatomy is unavoidable, which imposes a potential safety risk, including vessel wall damage or even perforation. This paper presents an approach for body contact estimation of continuum robots with tension-profile sensing of actuation fibers. First, tension-sensing optical fibers with multiple inscribed fiber Bragg grating (FBG) sensors are used for both actuation and in-situ sensing of the continuum robot. Second, a beam theory-based mechanical model considering segmental differences, multiple fiber interactions and external force interactions is established, followed by robust estimation of contact positions and forces. Finally, detailed simulations are conducted to validate the accuracy and effectiveness of the proposed method. Experiments on a notched continuum robot are carried out, and the results show that the proposed approach can effectively recover in-situ segmental actuation forces without the need of explicit modeling of the friction between the fibers and guiding channels. The method enables the estimation of the number of contact points, as well as contact positions and contact forces along the body of the continuum robot.

Experimental validation of saliency maps for understanding deep neural networks for weld penetration prediction

Multi-label learning based deep transfer neural network for facial attribute classification

Current electrophysiology signal recording and mapping systems have limited dynamic range (DR) and bandwidth, which causes loss of valuable information during acquisition of cardiac signals. We evaluated a novel advanced signal processing platform with the objective to obtain and assess additional information of clinical importance. Over 10 canines, we compared intracardiac recordings within all cardiac chambers, in various rhythms, in pacing and during radiofrequency (RF) ablation across two platforms; a conventional system and the PURE EP™ [(PEP); Bio Sig Technologies, Inc., Los Angeles, CA, USA]. Recording cardiac signals with varying amplitudes were consistently and reproducibly observed, without loss of detail or introduction of artefact. Further the amplitude of current of injury (COI) on the unipolar signals correlated with the instantaneous contact force (CF) recorded on the sensing catheter in all the animals (r2 = 0.94 in ventricle). The maximum change in the unipolar COI correlated with the change in local electrogram amplitude during non-irrigated RF ablation (r2 = 0.61 in atrium). Reduction in artefact attributable to pacing (20 sites) and noise during ablation (48 sites) was present on the PEP system. Within the PEP system, simultaneous display of identical signals, filtered differently, aided the visualization of discrete conduction tissue signals. Compared to current system, the PEP system provided incremental information including identifying conduction tissue signals, estimates of CF and a surrogate for lesion formation. This novel signal processing platform with increased DR and minimal front-end filtering may be useful in clinical practice.

In this study, the deep multi-layered group method of data handling (GMDH)-type neural network algorithm using revised heuristic self-organization method is proposed and applied to medical image diagnosis of liver cancer. The deep GMDH-type neural network can automatically organize the deep neural network architecture which has many hidden layers. The structural parameters such as the number of hidden layers, the number of neurons in hidden layers and useful input variables are automatically selected to minimize prediction error criterion defined as Akaike’s information criterion (AIC) or prediction sum of squares (PSS). The architecture of the deep neural network is automatically organized using the revised heuristic self-organization method which is a type of the evolutionary computation. This new neural network algorithm is applied to the medical image diagnosis of the liver cancer and the recognition results are compared with the conventional 3-layered sigmoid function neural network.

Survey on Deep Neural Networks in Speech and Vision Systems

We study the eigenvalue distribution of random matrices pertinent to the analysis of deep neural networks. The matrices resemble the product of the sample covariance matrices, however, an important difference is that the analog of the population covariance matrix is now a function of random data matrices (synaptic weight matrices in the deep neural network terminology). The problem has been treated in recent work [J. Pennington, S. Schoenholz and S. Ganguli, The emergence of spectral universality in deep networks, Proc. Mach. Learn. Res. 84 (2018) 1924–1932, arXiv:1802.09979] by using the techniques of free probability theory. Since, however, free probability theory deals with population covariance matrices which are independent of the data matrices, its applicability in this case has to be justified. The justification has been given in [L. Pastur, On random matrices arising in deep neural networks: Gaussian case, Pure Appl. Funct. Anal. (2020), in press, arXiv:2001.06188] for Gaussian data matrices with independent entries, a standard analytical model of free probability, by using a version of the techniques of random matrix theory. In this paper, we use another version of the techniques to extend the results of [L. Pastur, On random matrices arising in deep neural networks: Gaussian case, Pure Appl. Funct. Anal. (2020), in press, arXiv:2001.06188] to the case where the entries of the data matrices are just independent identically distributed random variables with zero mean and finite fourth moment. This, in particular, justifies the mean field approximation in the infinite width limit for the deep untrained neural networks and the property of the macroscopic universality of random matrix theory in this case.

Self-mixing interferometry (SMI) is superior to other laser interferometry methods due to its simplicity and compactness. However, SMI signals are often complex and difficult to process due to interference in the form of variations in the effective reflectivity of the target, noisy signals, complex signal shapes, and other dependencies. Deep neural networks have been a very popular area of research in computer artificial intelligence in recent years, allowing more implicit features to be uncovered than traditional shallow machine learning. It has been shown that convolutional and back-propagation neural networks can be used for SMI signal processing. There are also studies that have used machine learning genetic algorithms for absolute distance measurement. Based on the above, this study used convolutional neural networks to form a deep neural network for absolute distance measurement based on SMI technology. We first trained the deep convolutional neural network at different feedback strengths. The results of the Convolutional Neural Network (CNN) model showed a coefficient of determination of 0.9987. which is consistent with the required model. The trained network was then used to estimate absolute distances with and without the addition of noise. The comparison proves that the proposed method is noise-proof and has high adaptability for measurements under different conditions.

Gait recognition is a noncontact biometric procedure that determines the identity or health status of a person by analyzing his or her walking posture and habits, including skeletal and joint movements. The most remarkable feature of this method is the possibility of conducting recognition without demanding much cooperation from participants. Therefore, this recognition technique has attracted much attention from scholars. Additionally, because of the rapid development of graphics processing unit technology, related hardware and computation performance, the applications of deep-learning technology are considerably enhanced. The objective of this study was to apply a deep neural network (DNN), which employs deep-learning technology, to achieve gait-based automatic pedestrian detection and recognition. In contrast to using wearable devices to precisely capture skeletal and joint movements, pedestrian color-image sequences were used as input in this study. Subsequently, a pretraining convolutional neural network (CNN) was employed to capture pedestrian location and extract pedestrian dense optical flow to serve as concrete low-level feature inputs. Then, a finely-tuned DNN based on the wide residual network was employed to extract high-level abstract features. In addition, to overcome the difficulty of obtaining local temporal features by using a 2D CNN, part of the 3D convolutional structure was introduced into the CNN. This design enabled use of limited memory to acquire more effective features and enhance the DNN performance. The experimental results show that the proposed method has exceptional performance for pedestrian detection and recognition.