Adaptive Recognition Research Articles

Research on lithium battery sorting method based on image adaptive recognition of feature points

Research on lithium battery sorting method based on image adaptive recognition of feature points

Adaptive Recognition of Open-Set Voiceprint Based on Deep Learning Combined with MSM-SVDD Method

Adaptive Recognition of Open-Set Voiceprint Based on Deep Learning Combined with MSM-SVDD Method

Research on lithium battery sorting method based on image adaptive recognition of feature points

When sorting batteries in factories, fixed feature points in a discharge voltage platform are usually used as a sorting basis with expert experience, causing poor accuracy in battery sorting. This study proposes a method based on feature point image adaptive recognition (FPIAR) and combined a semi-supervised fuzzy C-means (SFCM) clustering algorithm to achieve battery sorting. Accordingly, FPIAR obtained the multi-dimensional feature coordinate dataset by adaptively identifying the discharge voltage curve of each battery. The centre point of the accumulation of coordinates in the dataset was found by mean-shift algorithm and was used as the feature point of battery sorting voltage. In addition, the SFCM clustering algorithm was used to process the feature point of the sorting voltage to sort the batteries into groups. The simulation results and tests indicated that the SFCM clustering algorithm based on FPIAR processing can improve the accuracy of battery sorting.

Adaptive demodulation by deep-learning-based identification of fractional orbital angular momentum modes with structural distortion due to atmospheric turbulence

Since the great success of optical communications utilizing orbital angular momentum (OAM), increasing the number of addressable spatial modes in the given physical resources has always been an important yet challenging problem. The recent improvement in measurement resolution through deep-learning techniques has demonstrated the possibility of high-capacity free-space optical communications based on fractional OAM modes. However, due to a tiny gap between adjacent modes, such systems are highly susceptible to external perturbations such as atmospheric turbulence (AT). Here, we propose an AT adaptive neural network (ATANN) and study high-resolution recognition of fractional OAM modes in the presence of turbulence. We perform simulations of fractional OAM beams propagating through a 1-km optical turbulence channel and analyze the effects of turbulence strength, OAM mode interval, and signal noise on the recognition performance of the ATANN. The recognition of multiplexed fractional modes is also investigated to demonstrate the feasibility of high-dimensional data transmission in the proposed deep-learning-based system. Our results show that the proposed model can predict transmitted modes with high accuracy and high resolution despite the collapse of structured fields due to AT and provide stable performance over a wide SNR range.

Hybrid two-stream dynamic CNN for view adaptive human action recognition using ensemble learning

Hybrid two-stream dynamic CNN for view adaptive human action recognition using ensemble learning

RRSARNet: A Novel Network for Radar Radio Sources Adaptive Recognition

Radar radio source (RRS) recognition plays an important role in the fields of military electronic support systems (ESM) and civilian autonomous driving. The rapid development of machine learning technology, especially deep learning, has effectively and efficiently improved RRS intelligent recognition performances when operating in the increasingly complex electromagnetic environment. However, the data sampling limitation and computation cost are still severe challenges in real RRS recognition scenarios. In this paper, we propose a novel network based on meta-transfer learning, called RRSARNet, to achieve effective adaptive RRS recognition in the context of low signal-to-noise ratio (SNR). First, by using the short-time Fourier transform, a six-type small samples RRS simulation dataset with different SNR levels is constructed. Then, a novel RRSARNet, based on metric learning, is proposed, which consists of a four-layer embedding module and a four-layer relational module. Finally, the RRS dataset is divided into training, supporting and testing subsets, which are used to train and test the RRSARNet in a meta-transfer learning method. Experiments on the RRS dataset show that the proposed RRSARNet can achieve an overall accuracy (OA) above 96% and 99% when the SNR is above −15 dB and −10 dB, respectively. Even when the SNR is -30 dB, OA can reach more than 70%. For 5-way 1-shot and 5-way 5-shot experiments, the inference time of an image is about 0.043 and 0.140 milliseconds, respectively. Besides, experiments on the RRS simulation dataset and the two benchmark datasets, the RRSARNet performs better or more competitive than many existing state-of-the-art technologies in terms of recognition accuracy.

Technology Recommendations for an Innovative Agricultural Robot Design Based on Technology Knowledge Graphs

The process of agricultural robot design is a complex system requiring the cooperation and integration of agricultural, machinery, automation, and information technology. These demands create great challenges for the innovative design of agricultural robots. Meanwhile, more than 95% of the latest inventions and creations in the world are recorded in the patent literature. In order to make effective use of the information and data resources of patents, shorten the design cycle, and provide knowledge for the designers, according to the operation’s objectives, an agricultural robot technology knowledge graph (TKG) was established for innovative designs. By analyzing the patent information, a patent IPC co-classification network (IPCNet) for adaptive design process recognition was put forward to meet the requirements of the different operation objectives and operation links. Through the extraction of the technology keywords and efficacy keywords, based on the word co-occurrence network (WCONet), a technology–efficacy map (TEM) was constructed. Through the integration of the adaptive design process and the TEM, the agricultural robot design TKG was constructed for determining technological recommendations for agricultural robot design. The case of the citrus picking robot design was realized to implement the design process. With the technology recommendation results, the moving system, body, and end-effector for the citrus picking robot were designed to verify the results of the recommendation.

Human intestinal and circulating invariant natural killer T cells are cytotoxic against colorectal cancer cells via the perforin-granzyme pathway.

Invariant natural killer T (iNKT) cells are lipid‐specific T lymphocytes endowed with cytotoxic activities and are thus considered important in antitumor immunity. While several studies have demonstrated iNKT cell cytotoxicity against different tumors, very little is known about their cell‐killing activities in human colorectal cancer (CRC). Our aim was to assess whether human iNKT cells are cytotoxic against colon cancer cells and the mechanisms underlying this activity. For this purpose, we generated stable iNKT cell lines from peripheral blood and colon specimens and used NK‐92 and peripheral blood natural killer cells as cell‐mediated cytotoxicity controls. In vitro cytotoxicity was assessed using a panel of well‐characterized human CRC cell lines, and the cellular requirements for iNKT cell cytotoxic functions were evaluated. We demonstrated that both intestinal and circulating iNKT cells were cytotoxic against the entire panel of CRC lines, as well as against freshly isolated patient‐derived colonic epithelial cancer cells. Perforin and/or granzyme inhibition impaired iNKT cell cytotoxicity, whereas T‐cell receptor (TCR) signaling was a less stringent requirement for efficient killing. This study is the first evidence of tissue‐derived iNKT cell cytotoxic activity in humans, as it shows that iNKT cells depend on the perforin–granzyme pathway and both adaptive and innate signal recognition for proper elimination of colon cancer cells.

Predicting children's audiovisual speech recognition thresholds based on unimodal performance

This study examined the extent to which individual differences in children’s audiovisual speech recognition thresholds are explained by their performance in unimodal auditory and unimodal visual conditions. Auditory-only and audiovisual speech recognition thresholds were measured in fifty children between 7 and 9 years of age using an adaptive closed-set speech recognition task in a two-talker speech masker. Children heard color-number combinations and responded by choosing the color-number combination on a touch-screen monitor. Visual-only speech recognition accuracy scores were measured using the same stimuli and closed-set response format. Correlations and linear modeling were used to examine the extent to which visual-only speech recognition accuracy, auditory-only speech recognition thresholds, and age predict audiovisual speech recognition thresholds. Audiovisual speech recognition thresholds were positively correlated with auditory-only speech recognition thresholds and negatively correlated with visual-only accuracy. A linear model explained 38% of variance in audiovisual thresholds and indicated that children with better visual-only accuracy had lower audiovisual thresholds. Effects of auditory-only accuracy and age were non-significant. These results suggest that lipreading ability contributes to individual differences in 7- to 9-year-old children’s audiovisual recognition of color-number stimuli in a two-talker speech masker.

How do children combine visual speech with other auditory grouping cues?

Adults use visual speech to help perceptually group talkers into separate objects and to attend to the target talker. As a result, they benefit more from visual speech in the presence of a two-talker speech masker than in speech-spectrum noise and benefit less from visual speech when other cues are available to support auditory grouping. Recent studies demonstrated that children’s visual speech benefit was similar for a noise masker and a two-talker speech masker, suggesting that children may not use visual speech as a grouping cue. The current study further examined whether children use visual speech as a grouping cue by testing whether 40 seven- to nine-year-old children’s visual speech benefit decreased in the presence of other cues known to promote auditory grouping in children, namely carrier phrases and perceived spatial separation. Children completed an adaptive closed-set speech recognition task in a two-talker speech masker, in auditory-only and audiovisual conditions, and in the presence or absence of one of the auditory grouping cues. Preliminary analyses suggest that children’s visual speech benefit did not decrease when other auditory grouping cues were present. These results provide converging evidence that children in this age range may not use visual speech as a grouping cue.

Adaptive Recognition of Motion Posture in Sports Video Based on Evolution Equation

In the field of sports, the formulation of existing training plans mainly relies on the manual observation and personal experience of coaches. This method is inevitably subjective. The application of artificial intelligence technology to the training of athletes to recognize athletes’ posture can help coaches assist in decision-making and greatly enhance athletes’ competitive ability. The human body movements embodied in sports are more complicated, and the accurate recognition of sports postures plays an active and important role in sports competitions and training. In this paper, inertial sensor technology is applied to attitude recognition in motion. First, in order to improve the accuracy of attitude calculation and reduce the noise interference in the preparation process, this article uses differential evolution algorithm to apply attitude calculation to realize multisensor data fusion. Secondly, a two-level neural network intelligent motion gesture recognition algorithm is proposed. The two-level neural network intelligent recognition algorithm effectively recognizes similar actions by splitting the traditional single-level neural network into two-level neural networks. Experiments show that the experimental method designed in this article for the posture in motion can obtain the motion information of the examinee in real time, realize the accurate extraction of individual motion data, and complete the recognition of the motion posture. The average accuracy rate can reach 98.85%. There is a certain practical value in gesture recognition.

Applying incremental Deep Neural Networks-based posture recognition model for ergonomics risk assessment in construction

Monitoring and assessing awkward postures is a proactive approach for Musculoskeletal Disorders (MSDs) prevention in construction. Machine Learning models have shown promising results when used in recognition of workers’ posture from Wearable Sensors. However, there is a need to further investigate: i) how to enable Incremental Learning, where trained recognition models continuously learn new postures from incoming subjects while controlling the forgetting of learned postures; ii) the validity of ergonomics risk assessment with recognized postures. The research discussed in this paper seeks to address this need through an adaptive posture recognition model– the incremental Convolutional Long Short-Term Memory (CLN) model. The paper discusses the methodology used to develop and validate this model’s use as an effective Incremental Learning strategy. The evaluation was based on real construction workers’ natural postures during their daily tasks. The CLN model with “shallow” (up to two) convolutional layers achieved high recognition performance (Macro F1 Score) under personalized (0.87) and generalized (0.84) modeling. Generalized CLN model, with one convolutional layer, using the “Many-to-One” Incremental Learning scheme can potentially balance the performance of adaptation and controlling forgetting. Applying the ergonomics rules on recognized and ground truth postures yielded comparable risk assessment results. These findings support that the proposed incremental Deep Neural Networks model has a high potential for adaptive posture recognition. They can be deployed alongside ergonomics rules for effective MSDs risk assessment.

Eye Movement Prediction Based on Adaptive BP Neural Network

This paper uses adaptive BP neural networks to conduct an in-depth examination of eye movements during reading and to predict reading effects. An important component for the implementation of visual tracking systems is the correct detection of eye movement using the actual data or real-world datasets. We propose the identification of three typical types of eye movements, namely, gaze, leap, and smooth navigation, using an adaptive BP neural network-based recognition algorithm for eye movement. This study assesses the BP neural network algorithm using the eye movement tracking sensors. For the experimental environment, four types of eye movement signals were acquired from 10 subjects to perform preliminary processing of the acquired signals. The experimental results demonstrate that the recognition rate of the algorithm provided in this paper can reach up to 97%, which is superior to the commonly used CNN algorithm.

Self-mediated positive selection of T cells sets an obstacle to the recognition of nonself

Adaptive immune recognition is mediated by the binding of peptide-human leukocyte antigen complexes by T cells. Positive selection of T cells in the thymus is a fundamental step in the generation of a responding T cell repertoire: only those T cells survive that recognize human peptides presented on the surface of cortical thymic epithelial cells. We propose that while this step is essential for optimal immune function, the process results in a defective T cell repertoire because it is mediated by self-peptides. To test our hypothesis, we focused on amino acid motifs of peptides in contact with T cell receptors. We found that motifs rarely or not found in the human proteome are unlikely to be recognized by the immune system just like the ones that are not expressed in cortical thymic epithelial cells or not presented on their surface. Peptides carrying such motifs were especially dissimilar to human proteins. Importantly, we present our main findings on two independent T cell activation datasets and directly demonstrate the absence of naïve T cells in the repertoire of healthy individuals. We also show that T cell cross-reactivity is unable to compensate for the absence of positively selected T cells. Additionally, we show that the proposed mechanism could influence the risk for different infectious diseases. In sum, our results suggest a side effect of T cell positive selection, which could explain the nonresponsiveness to many nonself peptides and could improve the understanding of adaptive immune recognition.

Nonstationary signal extraction based on BatOMP sparse decomposition technique

Sparse decomposition technique is a new method for nonstationary signal extraction in a noise background. To solve the problem of accuracy and efficiency exclusive in sparse decomposition, the bat algorithm combined with Orthogonal Matching Pursuits (BatOMP) was proposed to improve sparse decomposition, which can realize adaptive recognition and extraction of nonstationary signal containing random noise. Two general atoms were designed for typical signals, and dictionary training method based on correlation detection and Hilbert transform was developed. The sparse decomposition was turned into an optimizing problem by introducing bat algorithm with optimized fitness function. By contrast with several relevant methods, it was indicated that BatOMP can improve convergence speed and extraction accuracy efficiently as well as decrease the hardware requirement, which is cost effective and helps broadening the applications.

Application of adaptive technology-based speech recognition system in 600 pathological grossing process

Objective: To establish a speech recognition system based on adaptive technology and to evaluate its value in pathological grossing processes. Methods: A total of 600 tissue specimens were collected at the Zhejiang Provincial People's Hospital affiliated to Hangzhou Medical College between October 1, 2020 and December 31, 2020. A speech recognition system based on adaptive technology was used in the pathological grossing processes, and the pathological examination reports were generated and extracted. Results: The speech recognition system based on adaptive technology showed a good recognition rate (overall recognition rate = 77.87%) and helped achieve rapid input and output of pathological examination data. Conclusions: The speech recognition system can reduce the labor costs, improve the work efficiency of pathologists and increase the quality of medical services, which may be valuable for building next-generation smart hospitals.

Fate mapping of single NK cells identifies a type 1 innate lymphoid-like lineage that bridges innate and adaptive recognition of viral infection

SummaryUpon viral infection, natural killer (NK) cells expressing certain germline-encoded receptors are selected, expanded, and maintained in an adaptive-like manner. Currently, these are thought to differentiate along a common pathway. However, by fate mapping of single NK cells upon murine cytomegalovirus (MCMV) infection, we identified two distinct NK cell lineages that contributed to adaptive-like responses. One was equivalent to conventional NK (cNK) cells while the other was transcriptionally similar to type 1 innate lymphoid cells (ILC1s). ILC1-like NK cells showed splenic residency and strong cytokine production but also recognized and killed MCMV-infected cells, guided by activating receptor Ly49H. Moreover, they induced clustering of conventional type 1 dendritic cells and facilitated antigen-specific T cell priming early during MCMV infection, which depended on Ly49H and the NK cell-intrinsic expression of transcription factor Batf3. Thereby, ILC1-like NK cells bridge innate and adaptive viral recognition and unite critical features of cNK cells and ILC1s.

Multiperson Interactive Activity Recognition Based on Interaction Relation Model

Multiperson activity recognition is a pivotal branch as well as a challenging topic of human action recognition research. This paper adopts a hybrid learning model to the spatio-temporal relationship and occlusion relationship among multiple people. Initially, this paper builds up an active multiperson interaction relationship estimation framework model to capture interpersonal spatio-temporal relation. This model incorporates the interaction relationship estimation framework with the multiperson relationship network. On this ground, it automatically learns from the human-computer interaction dataset in an end-to-end manner and performs reasoning with standard matrix operations. Secondly, this paper proposed an adaptive occlusion state behavior recognition method derived from the semantic knowledge model to ravel out the concern of occlusion and self-occlusion in human action recognition. Then, Petri Nets are used to recognize multiperson interactive actions. This model has been through extensive experiments on the TV interaction dataset, Vlog dataset, AVA dataset, and MLB-YouTube dataset, experimental results have proved that the recognition performance of this model is superior than the other available models. This paper prospects and summarizes the estimation framework of the interaction relationship and occlusion semantic-knowledge relationship. Experimental results suggest that the proposed method in the paper could capture the discriminative relation information for multiperson interactive activity recognition, which further validates the efficiency of the hybrid learning model.

An adaptive recognition method for take-off action images of back-style high jump based on feature extraction

In the process of human movement, the behavioral characteristics of human body are complex and diverse. Thus, it is difficult to use accurate shape to represent the uncertainty of human behavior, which brings difficulties for 3D modeling in the state of motion. To solve this problem, an adaptive image recognition method based on feature extraction is proposed in this paper. Firstly, the image noise threshold was obtained by wavelet coefficient conversion. Then the wavelet coefficient matrix was constructed on the basis of decomposed image multi-resolution, and the image noise was removed according to the threshold. Based on this, Gaussian function in Canny method is introduced to obtain the edge coordinates of moving targets by convolution operation. After that, the image rotation and translation matrices were resolved according to the Murkowski distance to obtain the motion feature vectors. Finally, the hidden Markov model was constructed to segment the background and the target according to the feature sequence and the target edge coordinates, so as to realize the adaptive recognition of the take-off action images of the back-style high jump. Experimental results showed that this method had the advantages of high denoising effect, accurate feature extraction, error free edge division and accurate recognition result.

Viral Control of Glioblastoma.

Glioblastoma multiforme (GBM) is a universally lethal cancer of the central nervous system. Patients with GBM have a median survival of 14 months and a 5-year survival of less than 5%, a grim statistic that has remained unchanged over the last 50 years. GBM is intransigent for a variety of reasons. The immune system has a difficult time mounting a response against glioblastomas because they reside in the brain (an immunologically dampened compartment) and generate few neoantigens relative to other cancers. Glioblastomas inhabit the brain like sand in the grass and display a high degree of intra- and inter-tumoral heterogeneity, impeding efforts to therapeutically target a single pathway. Of all potential therapeutic strategies to date, virotherapy offers the greatest chance of counteracting each of the obstacles mounted by GBM. Virotherapy can xenogenize a tumor that is deft at behaving like “self”, triggering adaptive immune recognition in an otherwise immunologically quiet compartment. Viruses can also directly lyse tumor cells, creating damage and further stimulating secondary immune reactions that are detrimental to tumor growth. In this review, we summarize the basic immune mechanisms underpinning GBM immune evasion and the recent successes achieved using virotherapies.