Deep Learning-based Convolutional Neural Network Research Articles

3-D Tactile-Based Object Recognition for Robot Hands Using Force-Sensitive and Bend Sensor Arrays

Tactile sensing is a particularly important and challenging task for a modern robot to safely manipulate objects, interact with humans in a shared space, and provide various services. This paper presents a 3D tactile glove for robots with the combination of a piezoresistive-based force sensor array (412 sensors) covering the full hand and a resistive bend sensor array (5 sensors) on the back of five fingers. Deep learning-based CNN (Convolutional Neural Network) methods using the designed tactile glove are proposed for object recognition. In the experiment for recognizing 15 objects with a dexterous robot hand, an average classification accuracy of 93.67% has been achieved. Comparison experiments with three other typical classifiers (the Quadratic SVM, Weighted KNN, and Bagged Trees) and our CNN methods show an average recognition accuracy of 91.67% with the 3D tactile glove, revealing an accuracy improvement of 4.17% over only using the force sensor array. We further apply our 3D tactile glove and the multi-modal CNN to identify three other objects and demonstrate their generalization ability of tactile object recognition with an average success accuracy of 78.33%. The proposed 3D tactile glove can be further used in human-robot interactions, the design of prosthetics and humanoid robots, and for improving the intelligence level in brain-computer collaborative systems.

A Deep-Learning Algorithm to Predict Short-Term Progression to Geographic Atrophy on Spectral-Domain Optical Coherence Tomography

The identification of patients at risk of progressing from intermediate age-related macular degeneration (iAMD) to geographic atrophy (GA) is essential for clinical trials aimed at preventing disease progression. DeepGAze is a fully automated and accurate convolutional neural network-based deep learning algorithm for predicting progression from iAMD to GA within 1 year from spectral-domain optical coherence tomography (SD-OCT) scans. To develop a deep-learning algorithm based on volumetric SD-OCT scans to predict the progression from iAMD to GA during the year following the scan. This retrospective cohort study included participants with iAMD at baseline and who either progressed or did not progress to GA within the subsequent 13 months. Participants were included from centers in 4 US states. Data set 1 included patients from the Age-Related Eye Disease Study 2 AREDS2 (Ancillary Spectral-Domain Optical Coherence Tomography) A2A study (July 2008 to August 2015). Data sets 2 and 3 included patients with imaging taken in routine clinical care at a tertiary referral center and associated satellites between January 2013 and January 2023. The stored imaging data were retrieved for the purpose of this study from July 1, 2022, to February 1, 2023. Data were analyzed from May 2021 to July 2023. A position-aware convolutional neural network with proactive pseudointervention was trained and cross-validated on Bioptigen SD-OCT volumes (data set 1) and validated on 2 external data sets comprising Heidelberg Spectralis SD-OCT scans (data sets 2 and 3). Prediction of progression to GA within 13 months was evaluated with area under the receiver-operator characteristic curves (AUROC) as well as area under the precision-recall curve (AUPRC), sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. The study included a total of 417 patients: 316 in data set 1 (mean [SD] age, 74 [8]; 185 [59%] female), 53 in data set 2, (mean [SD] age, 83 [8]; 32 [60%] female), and 48 in data set 3 (mean [SD] age, 81 [8]; 32 [67%] female). The AUROC for prediction of progression from iAMD to GA within 1 year was 0.94 (95% CI, 0.92-0.95; AUPRC, 0.90 [95% CI, 0.85-0.95]; sensitivity, 0.88 [95% CI, 0.84-0.92]; specificity, 0.90 [95% CI, 0.87-0.92]) for data set 1. The addition of expert-annotated SD-OCT features to the model resulted in no improvement compared to the fully autonomous model (AUROC, 0.95; 95% CI, 0.92-0.95; P = .19). On an independent validation data set (data set 2), the model predicted progression to GA with an AUROC of 0.94 (95% CI, 0.91-0.96; AUPRC, 0.92 [0.89-0.94]; sensitivity, 0.91 [95% CI, 0.74-0.98]; specificity, 0.80 [95% CI, 0.63-0.91]). At a high-specificity operating point, simulated clinical trial recruitment was enriched for patients progressing to GA within 1 year by 8.3- to 20.7-fold (data sets 2 and 3). The fully automated, position-aware deep-learning algorithm assessed in this study successfully predicted progression from iAMD to GA over a clinically meaningful time frame. The ability to predict imminent GA progression could facilitate clinical trials aimed at preventing the condition and could guide clinical decision-making regarding screening frequency or treatment initiation.

Structural and functional connectome relationships in early childhood

There is strong evidence that the functional connectome is highly related to the white matter connectome in older children and adults, though little is known about structure-function relationships in early childhood. We investigated the development of cortical structure-function coupling in children longitudinally scanned at 1, 2, 4, and 6 years of age (N = 360) and in a comparison sample of adults (N = 89). We also applied a novel graph convolutional neural network-based deep learning model with a new loss function to better capture inter-subject heterogeneity and predict an individual’s functional connectivity from the corresponding structural connectivity. We found regional patterns of structure-function coupling in early childhood that were consistent with adult patterns. In addition, our deep learning model improved the prediction of individual functional connectivity from its structural counterpart compared to existing models.

PAIP 2020: Microsatellite instability prediction in colorectal cancer.

Microsatellite instability (MSI) refers to alterations in the length of simple repetitive genomic sequences. MSI status serves as a prognostic and predictive factor in colorectal cancer. The MSI-high status is a good prognostic factor in stage II/III cancer, and predicts a lack of benefit to adjuvant fluorouracil chemotherapy in stage II cancer but a good response to immunotherapy in stage IV cancer. Therefore, determining MSI status in patients with colorectal cancer is important for identifying the appropriate treatment protocol. In the Pathology Artificial Intelligence Platform (PAIP) 2020 challenge, artificial intelligence researchers were invited to predict MSI status based on colorectal cancer slide images. Participants were required to perform two tasks. The primary task was to classify a given slide image as belonging to either the MSI-high or the microsatellite-stable group. The second task was tumor area segmentation to avoid ties with the main task. A total of 210 of the 495 participants enrolled in the challenge downloaded the images, and 23 teams submitted their final results. Seven teams from the top 10 participants agreed to disclose their algorithms, most of which were convolutional neural network-based deep learning models, such as EfficientNet and UNet. The top-ranked system achieved the highest F1 score (0.9231). This paper summarizes the various methods used in the PAIP 2020 challenge. This paper supports the effectiveness of digital pathology for identifying the relationship between colorectal cancer and the MSI characteristics.

Enhancing heart disease diagnosis through precise heart sound including infrasound

Recent research has been pushing the frontier of diagnosing cardiovascular diseases, which have a high mortality rate globally. The need for swift responses to acute diseases has underscored the importance of continuous health monitoring and the development of monitoring diagnostic technology. Our work focused on an overlooked aspect—the microphones used to measure heart sounds. Existing ones, particularly capacitive MEMS types, measure only heart sounds primarily within the audible frequency range, potentially limiting the diagnostic accuracy. To overcome these limitations, we used microphones capable of measuring down to infrasound areas. We also designed an experimental environment to capture the precise heart sounds. The crux of our approach was using the full information of heart sounds through infrasound measurements. A convolutional neural network-based deep learning model was employed for algorithm development and validated using the PhysioNet 2016 CinC’s open-source heart sound data. We conducted an experiment on the impact of low-frequency components on heart sound diagnosis using a new infrasound dataset measured at Seoul St. Mary's Hospital of the Catholic University of Korea. Accuracy and sensitivity, respectively, improved on average by 2% and 4%, showing that datasets including low-frequency components generally performed better.

A Comparative Study of Object Detection using YOLO and SSD Algorithms

One of the most significant and difficult areas of computer vision is object detection, which is used in many aspects of daily life, including security monitoring, autonomous vehicle operation, and so on, to find instances of semantic objects belonging to a particular class. In this survey, we first look at the current approaches of typical detection models and discuss the benchmark datasets in order to comprehensively and deeply comprehend the major development status of the object detection pipeline. This project explains how convolutional neural network-based deep learning techniques are used for object detection. In this study, deep learning methods for cutting-edge object identification systems are evaluated. Any kind of visual medium is represented by a computer as a collection of numerical numbers. In order to inspect the contents of images, they need image processing techniques. In order to determine which is the quickest and most effective object detection algorithm, this research analyzes two popular algorithms: Single Shot Detection (SSD) and You Only Look Once (YOLO). The COCO (Common Object in Context) dataset is used in this comparative research to assess the effectiveness of the two algorithms and to analyze their strengths and weaknesses in terms of factors like accuracy, precision, and speed. Based on the findings analysis, it can be said that YOLO-v3 outperforms SSD in the same testing scenario. Keywords- Object Detection, Deep learning, CNN, COCO, YOLO, SSD

Color image information transmission based on elliptic optical vortex array encoding/decoding.

A multichannel high-dimensional data encoding/decoding scheme based on composite elliptic optical vortex (EOV) arrays is proposed. By exploiting the rotation angle of the EOV, a 4 × 4 composite EOV array is used for high-dimensional data encoding. The conjugate symmetric extension Fourier computer-generated holography algorithm with controllable reconstruction focus is used to assign different reconstruction focus to the data of the three channels (R, G, and B) of the color image. Then, the data of the three channels is transmitted simultaneously by a single hologram to further improve the transmission efficiency. At the receiver, the initial information sequence is decoded by directly identifying the captured intensity patterns with a deep learning-based convolutional neural network. In the experiment, a 128 × 128-pixel color image is successfully transmitted, which confirms the feasibility of our proposed encoding/decoding scheme. This method has great potential for future high-capacity optical communications.

Search for two-neutrino double-beta decay of 136Xe to the excited state of 136Ba with the complete EXO-200 dataset* *EXO-200 is supported by DOE and NSF in the United States, NSERC in Canada, SNF in Switzerland, IBS in Korea, DFG in Germany, and CAS in China. EXO-200 data analysis and simulation uses resources of the National Energy Research Scientific Computing Center

A new search for two-neutrino double-beta (2νββ) decay of 136Xe to the excited state of 136Ba is performed with the full EXO-200 dataset. A deep learning-based convolutional neural network is used to discriminate signal from background events. Signal detection efficiency is increased relative to previous searches by EXO-200 by more than a factor of two. With the addition of the Phase II dataset taken with an upgraded detector, the median 90% confidence level half-life sensitivity of 2νββ decay to the state of 136Ba is yr using a total 136Xe exposure of 234.1 kg yr. No statistically significant evidence for 2νββ decay to the state is observed, leading to a lower limit of yr at 90% confidence level, improved by 70% relative to the current world's best constraint.

Deep learning for anomaly detection in wire-arc additive manufacturing

Wire-arc additive manufacturing (WAAM) is becoming the most important metal additive manufacturing process in many industries. In this paper, one of the common problems of irregularity in the metal deposition in WAAM has been addressed and solved using machine learning (ML). A deep learning-based convolutional neural network (CNN) was used to classify the two classes of deposited beads, i.e. ‘regular bead’ and ‘irregular bead’. A digital camera was installed with a WAAM setup to obtain the images of beads after deposition. A single layer of deposition was conducted on a substrate using aluminium 5356 alloy filler wire using robotic-controlled gas-metal arc welding (GMAW) setup. The performance of the ML model was validated using classification accuracy and processing time. The developed CNN model was checked with three types of proposed datasets. The dataset containing the training and testing ratio of 60:40 achieved an accuracy of 86.53% and 88.08% with 30 and 60 epochs respectively for testing. The proposed ML model was successful in anomaly detection in the deposited bead of WAAM and hence it helps in improving the quality of deposited layers and mechanical properties of fabricated parts.

GLDM: Geo-location prediction of twitter users with deep learning methods1

Social networks like Twitter are extremely popular and widely used, which has increased interest in studying the information posted there. One such analytical application is extracting location information of users for real-time monitoring of the objects and events of interest, such as political and social events, disease surveillance, natural calamities, and crime prevention. Identifying geographic location is a nontrivial task, as user profiles contain outdated and inaccurate location information. Furthermore, extracting geographical information from Arabic tweets is challenging since they contain many nonstandard data (dialects), complex structures, abbreviations, grammatical and spelling mistakes, etc. This study focuses on the localization of Saudi Arabian users who tweet in Arabic. This study proposes a convolutional neural network-based deep learning model to predict a Twitter user’s region-level location using user profiles, text texts, place attachments, and historical tweets. The model was evaluated empirically on a dataset of 95,739 tweets written in Arabic and produced by 4,331 users from Saudi Arabia cities. Regarding classification accuracy, the proposed CNN model outperformed machine learning classifiers such as NB, LR, and SVM with a 60% accuracy on the test set. This study is the first of its kind, aimed at localizing Saudi users based on their tweets.

Surveillance audio-based rainfall observation: An enhanced strategy for extreme rainfall observation

The increased frequency of extreme rainfall events (EREs) causing disastrous effects on society has become an indisputable fact in recent years. The weak performance of the current observation network in terms of accuracy and spatiotemporal resolution makes the development of new rainfall estimation techniques essential to reduce the impact of ERE-driven disasters. The rainfall sounds collected by widespread surveillance cameras provide an opportunity to produce high-resolution, all-weather rainfall data. However, the complex structure of the ground surface and the random background noise make building an effective surveillance audio-based ERE estimation system challenging. In this study, from the viewpoint of surveillance sound space, a 3D printer was used to create a shelter for the surveillance camera to define the underlying surface of falling raindrops artificially. Combining the knowledge of meteorology, micro-physics, and acoustics of rainfall, the shelter structure was designed to standardize the acoustical behavior while enhancing the consistency and specificity of raindrop sound, especially in complex scenarios such as those disturbed by different levels of wind. After that, convolutional neural network-based deep learning algorithms were used to classify ERE levels, and an audio-based ERE classification system was built. The experimental results show that the shelter facilitates audio-based rainfall representation; moreover, with the help of shelter, our proposed system achieved performance with about 93.4% accuracy in complex rainfall scenarios. Our study supports high-resolution rainfall data production on existing surveillance resources, developing a novel and reliable alternative for the perception of ERE and the calibration of observations from current rainfall networks.

Deep Learning Convolutional Neural Network Reconstruction and Radial k-Space Acquisition MR Technique for Enhanced Detection of Retropatellar Cartilage Lesions of the Knee Joint.

To assess diagnostic performance of standard radial k-space (PROPELLER) MRI sequences and compare with accelerated acquisitions combined with a deep learning-based convolutional neural network (DL-CNN) reconstruction for evaluation of the knee joint. Thirty-five patients undergoing MR imaging of the knee at 1.5 T were prospectively included. Two readers evaluated image quality and diagnostic confidence of standard and DL-CNN accelerated PROPELLER MR sequences using a four-point Likert scale. Pathological findings of bone, cartilage, cruciate and collateral ligaments, menisci, and joint space were analyzed. Inter-reader agreement (IRA) for image quality and diagnostic confidence was assessed using intraclass coefficients (ICC). Cohen's Kappa method was used for evaluation of IRA and consensus between sequences in assessing different structures. In addition, image quality was quantitatively evaluated by signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) measurements. Mean acquisition time of standard vs. DL-CNN sequences was 10 min 3 s vs. 4 min 45 s. DL-CNN sequences showed significantly superior image quality and diagnostic confidence compared to standard MR sequences. There was moderate and good IRA for assessment of image quality in standard and DL-CNN sequences with ICC of 0.524 and 0.830, respectively. Pathological findings of the knee joint could be equally well detected in both sequences (κ-value of 0.8). Retropatellar cartilage could be significantly better assessed on DL-CNN sequences. SNR and CNR was significantly higher for DL-CNN sequences (both p < 0.05). In MR imaging of the knee, DL-CNN sequences showed significantly higher image quality and diagnostic confidence compared to standard PROPELLER sequences, while reducing acquisition time substantially. Both sequences perform comparably in the detection of knee-joint pathologies, while DL-CNN sequences are superior for evaluation of retropatellar cartilage lesions.

A deep learning-based framework for multi-source precipitation fusion

Accurate quantitative precipitation estimation (QPE) is essential for various applications, including land surface modeling, flood forecasting, drought monitoring and prediction. In situ precipitation datasets, remote sensing-based estimations, and reanalysis products have heterogeneous uncertainty. Numerous models have been developed to merge precipitation estimations from different sources to improve the accuracy of QPE. However, many of these attempts are mainly focused on spatial or temporal correlations between various remote sensing sources and/or gauge data separately, and thus, the developed model cannot fully capture the inherent spatiotemporal dependencies that could potentially improve the precipitation estimations. In this study, we developed a general framework that can simultaneously merge and downscale multiple user-defined precipitation products by using rain gauge observations as target values. A novel deep learning-based convolutional neural network architecture, namely, the precipitation data fusion network (PDFN), that combines multiple layers of 3D-CNN and ConvLSTM was developed to fully exploit the spatial and temporal patterns of precipitation. This architecture benefits from techniques such as batch normalization, data augmentation schemes, and dropout layers to avoid overfitting and address skewed class proportions due to the highly imbalanced nature of the precipitation datasets. The results showed that the fused daily product remarkably improved the mean square error (MSE) and Pearson correlation coefficient (PCC) by 35% and 16%, respectively, compared to the best-performing product.

A lightweight model for efficient identification of plant diseases and pests based on deep learning.

Plant diseases and pests have always been major contributors to losses that occur in agriculture. Currently, the use of deep learning-based convolutional neural network models allows for the accurate identification of different types of plant diseases and pests. To enable more efficient identification of plant diseases and pests, we design a novel network architecture called Dise-Efficient based on the EfficientNetV2 model. Our experiments demonstrate that training this model using a dynamic learning rate decay strategy can improve the accuracy of plant disease and pest identification. Furthermore, to improve the model's generalization ability, transfer learning is incorporated into the training process. Experimental results indicate that the Dise-Efficient model boasts a compact size of 13.3 MB. After being trained using the dynamic learning rate decay strategy, the model achieves an accuracy of 99.80% on the Plant Village plant disease and pest dataset. Moreover, through transfer learning on the IP102 dataset, which represents real-world environmental conditions, the Dise-Efficient model achieves a recognition accuracy of 64.40% for plant disease and pest identification. In light of these results, the proposed Dise-Efficient model holds great potential as a valuable reference for the deployment of automatic plant disease and pest identification applications on mobile and embedded devices in the future.

Convolutional neural network-based deep learning model for air quality prediction in October city of Egypt

Purpose Modern human society has continuous advancements that have a negative impact on the quality of the air. Daily transportation, industrial and residential operations churn up dangerous contaminants in our surroundings. Addressing air pollution issues is critical for human health and ecosystems, particularly in developing countries such as Egypt. Excessive levels of pollutants have been linked to a variety of circulatory, respiratory and nervous illnesses. To this end, the purpose of this research paper is to forecast air pollution concentrations in Egypt based on time series analysis. Design/methodology/approach Deep learning models are leveraged to analyze air quality time series in the 6th of October City, Egypt. In this regard, convolutional neural network (CNN), long short-term memory network and multilayer perceptron neural network models are used to forecast the overall concentrations of sulfur dioxide (SO2) and particulate matter 10 µm in diameter (PM10). The models are trained and validated by using monthly data available from the Egyptian Environmental Affairs Agency between December 2014 and July 2020. The performance measures such as determination coefficient, root mean square error and mean absolute error are used to evaluate the outcomes of models. Findings The CNN model exhibits the best performance in terms of forecasting pollutant concentrations 3, 6, 9 and 12 months ahead. Finally, using data from December 2014 to July 2021, the CNN model is used to anticipate the pollutant concentrations 12 months ahead. In July 2022, the overall concentrations of SO2 and PM10 are expected to reach 10 and 127 µg/m3, respectively. The developed model could aid decision-makers, practitioners and local authorities in planning and implementing various interventions to mitigate their negative influences on the population and environment. Originality/value This research introduces the development of an efficient time-series model that can project the future concentrations of particulate and gaseous air pollutants in Egypt. This research study offers the first time application of deep learning models to forecast the air quality in Egypt. This research study examines the performance of machine learning approaches and deep learning techniques to forecast sulfur dioxide and particular matter concentrations using standard performance metrics.

Emotion classification for musical data using deep learning techniques

This research is done based on the identification and thorough analyzing musical data that is extracted by the various method. This extracted information can be utilized in the deep learning algorithm to identify the emotion, based on the hidden features of the dataset. Deep learning-based convolutional neural network (CNN) and long short-term memory-gated recurrent unit (LSTM-GRU) models were developed to predict the information from the musical information. The musical dataset is extracted using the fast Fourier transform (FFT) models. The three deep learning models were developed in this work the first model was based on the information of extracted information such as zero-crossing rate, and spectral roll-off. Another model was developed on the information of Mel frequencybased cepstral coefficient (MFCC) features, the deep and wide CNN algorithm with LSTM-GRU bidirectional model was developed. The third model was developed on the extracted information from Mel-spectrographs and untied these graphs based on two-dimensional (2D) data information to the 2D CNN model alongside LSTM models. Proposed model performance on the information from Mel-spectrographs is compared on the F1 score, precision, and classification report of the models. Which shows better accuracy with improved F1 and recall values as compared with existing approaches.

Species classification of brassica napus based on flowers, leaves, and packets using deep neural networks

Deep learning (DL) has gradually taken the lead as the most effective approach in the agricultural fields due to the early identification and classification of plant species and diseases for improving the quality of crop production because of recent technological breakthroughs, which have had a significant impact on agriculture. Plenty of complicated problems in farming, including species classification, plant disorder identification, yield approximation, and weather and soil moisture prediction, are made simple using deep neural networks. Thus, this proposed study aims to classify Brassica Napus (B. Napus) rapeseed species based on their most significant features, like flowers, leaves, and packets. The study has adopted two types of rapeseed such as B. Rapa and B. Alba. Five contemporary deep learning-based Convolutional Neural Network (CNN) models have also been assessed for distinguishing rapeseed species. These models are DenseNet201, VGG19, InceptionV3, Xception, and ResNet50. Initially, the researchers collected data from the agricultural field, and then image pre-processing is performed to create our dataset. After that, CNN models were applied to this dataset and enumerated the experimental data accordingly. Our DenseNet201 model successfully classified both species with the highest accuracy of 100% for flowers and 97% for both packets and leaves. A comprehensive analysis with companion studies confirmed the efficacy of our preferred paradigm for the near future. Nevertheless, future studies will compare these methodologies to data from a separate metabolomics dataset from comparable crops.

Reconstructing the flame in supersonic combustion flow by using the distribution of the wall pressure based on deep learning

Two models based on the deep learning-based convolutional neural network (CNN) and the re-parameterized convolutional neural network (RepCNN) were designed to reconstruct the flame in the combustor. Experiments were performed on a ground-pulse combustion wind tunnel at a fixed inlet Mach number of 2.5 and different pressures to inject hydrogen to obtain the relevant datasets. The results showed that both models could reconstruct the image of the flame in the combustor based on pressures of the upper and lower walls as well as the pressure at which hydrogen was injected. The average structural similarity index between the reconstructed image of the flame and its actual/original image was 0.9553, the average peak signal-to-noise ratio was 34.201, and the average correlation coefficient was 0.9819. The speed of reconstruction of the image using the RepCNN model improved by 40.7% at the cost of a slightly lower accuracy compared with the CNN model, and it took only 2.85 ms to reconstruct the image of a single flame. The lightweight feature of the RepCNN provides an important foundation for monitoring the model to reconstruct the image of the flame in real time. The work here simplifies requirements on the hardware for ground wind tunnel tests and provides a new idea for examining the characteristics of the flame in small combustors.

Early Diagnosis of Lung Cancer with Advanced ALCDC System and Deep Learning-Based CNN Algorithm

Abstract: This research paper presents a study on the use of an advanced ALCDC system, which utilizes a deep learning-based convolutional neural network (CNN) algorithm, for early detection of lung cancer. The purpose of this study is to evaluate the effectiveness of this system in detecting lung cancer at an early stage and to compare its performance with traditional detection methods.Lung cancer is the leading cause of cancer-related deaths worldwide, with a high mortality rate due to late diagnosis. Therefore, early detection of lung cancer is crucial for successful treatment and improving patient outcomes. Traditional methods of lung cancer detection, such as chest X-rays and CT scans, have limitations in terms of accuracy and efficiency. Recent advancements in AI and deep learning algorithms, such as CNNs, have shown promise in improving the accuracy and efficiency of lung cancer detection.In comparison to traditional detection methods, the ALCDC system showed a significant improvement in accuracy and efficiency. The system was able to detect lung nodules at an earlier stage, which is critical for successful treatment and improving patient outcomes.

Application of image recognition based on neural network algorithm in multimedia dance teaching

Neural network recognition algorithm is a deep learning-based convolutional neural network, which has been applied to target detection, image recognition, sound recognition and other fields. Although the recognition technology of convolutional neural network has been gradually stabilized, it is mainly implemented in single-machine serial mode. Therefore, there will be problems such as too long training time and insufficient memory capacity. According to the previous work, this paper has carried out related research and obtained some useful results. This paper focuses on the introduction of image recognition technology based on neural network algorithm. At the same time, this paper proposes a fast real-time image recognition algorithm, which has high recognition accuracy. In this paper, the existing algorithms mentioned in the paper are analyzed, the characteristics and shortcomings of all algorithms are summarized, and the appropriate combination of algorithms is selected for specific problems. On this basis, the acceleration method of artificial neural network is used to improve the algorithm of convolutional neural network image recognition technology. With the continuous development of education and the increase of the number of students year by year, the effect of classroom teaching becomes particularly important. Therefore, modern information technology is used to transfer the classroom teaching mode to the network and realize a new method of online remote online dance teaching. According to the actual situation of dance teaching, the advantages of multimedia dance teaching are analyzed. The software of multimedia dance teaching system is tested comprehensively, from the aspects of computer system memory occupancy rate and temperature, to test whether the activities of B/S mode multimedia dance teaching system meet the expected goals and requirements of each core module of multimedia dance teaching system.