Search

The detection of image similarity is critical to trademark (TM) legal registration and court judgment on infringement cases. Meanwhile, there are great challenges regarding the annotation of similar pairs and model generalization on rapidly growing data when deep learning is introduced into the task. The research idea of metric learning is naturally suited for the task where similarity of input is given instead of classification, but current methods are not targeted at the task and should be upgraded. To address these issues, loss-driven model training is introduced, and a hybrid-margin softmax (HMS) is proposed exactly based on the peculiarity of TM images. Two additive penalty margins are attached to the softmax to expand the decision boundary and develop greater tolerance for slight differences between similar TM images. With the HMS, a Siamese neural network (SNN) as the feature extractor is further penalized and the discrimination ability is improved. Experiments demonstrate that the detection model trained on HMS can make full use of small numbers of training data and has great discrimination ability on bigger quantities of test data. Meanwhile, the model can reach high performance with less depth of SNN. Extensive experiments indicate that the HMS-driven model trained completely on TM data generalized well on the face recognition (FR) task, which involves another type of image data. [ABSTRACT FROM AUTHOR]

Apple orchards, as an important center of economic activity in forestry special crops, can achieve yield prediction and automated harvesting by detecting and locating apples. Small apples, occlusion, dim lighting at night, blurriness, cluttered backgrounds, and other complex scenes significantly affect the automatic harvesting and yield estimation of apples. To address these issues, this study proposes an apple detection algorithm, "YOLOv5-ACS (Apple in Complex Scenes)", based on YOLOv5s. Firstly, the space-to-depth-conv module is introduced to avoid information loss, and a squeeze-and-excitation block is added in C3 to learn more important information. Secondly, the context augmentation module is incorporated to enrich the context information of the feature pyramid network. By combining the shallow features of the backbone P2, the low-level features of the object are retained. Finally, the addition of the context aggregation block and CoordConv aggregates the spatial context pixel by pixel, perceives the spatial information of the feature map, and enhances the semantic information and global perceptual ability of the object. We conducted comparative tests in various complex scenarios and validated the robustness of YOLOv5-ACS. The method achieved 98.3% and 74.3% for mAP@0.5 and mAP@0.5:0.95, respectively, demonstrating excellent detection capabilities. This paper creates a complex scene dataset of apples on trees and designs an improved model, which can provide accurate recognition and positioning for automatic harvesting robots to improve production efficiency. [ABSTRACT FROM AUTHOR]

Purpose: We developed and tested an automatic field‐in‐field (FIF) solution for whole‐brain radiotherapy (WBRT) planning that creates a homogeneous dose distribution by minimizing hotspots, resulting in clinically acceptable plans. Methods: A configurable auto‐planning algorithm was developed to automatically generate FIF WBRT plans independent of the treatment planning system. Configurable parameters include the definition of hotspots, target volume, maximum number of subfields, and minimum number of monitor units per field. This algorithm iteratively identifies a hotspot, creates two opposing subfields, calculates the dose, and optimizes the beam weight based on user‐configured constraints of dose‐volume histogram coverage and least‐squared cost functions. The algorithm was retrospectively tested on 17 whole‐brain patients. First, an in‐house landmark‐based automated beam aperture technique was used to generate the treatment fields and initial plans. Second, the FIF algorithm was employed to optimize the plans using physician‐defined goals of 99.9% of the brain volume receiving 100% of the prescription dose (30 Gy in 10 fractions) and a target hotspot definition of 107% of the prescription dose. The final auto‐optimized plans were assessed for clinical acceptability by an experienced radiation oncologist using a five‐point scale. Results: The FIF algorithm reduced the mean (± SD) plan hotspot percentage dose from 35.0 Gy (116.6%) ± 0.6 Gy (2.0%) to 32.6 Gy (108.8%) ± 0.4 Gy (1.2%). Also, it decreased the mean (± SD) hotspot V107% [cm3] from 959 ± 498 cm3 to 145 ± 224 cm3. On average, plans were produced in 16 min without any user intervention. Furthermore, 76.5% of the auto‐plans were clinically acceptable (needing no or minor stylistic edits), and all of them were clinically acceptable after minor clinically necessary edits. Conclusions: This algorithm successfully produced high‐quality WBRT plans and can improve treatment planning efficiency when incorporated into an automatic planning workflow. [ABSTRACT FROM AUTHOR]

As a non-destructive detection method, an electronic nose can be used to assess the freshness of meats by collecting and analyzing their odor information. Deep learning can automatically extract features and uncover potential patterns in data, minimizing the influence of subjective factors such as selecting features artificially. A transfer-learning-based model was proposed for the electronic nose to detect the freshness of chicken breasts in this study. First, a 3D-printed electronic nose system is used to collect the odor data from chicken breast samples stored at 4 °C for 1–7 d. Then, three conversion to images methods are used to feed the recorded time series data into the convolutional neural network. Finally, the pre-trained AlexNet, GoogLeNet, and ResNet models are retrained in the last three layers while being compared to classic machine learning methods such as K Nearest Neighbors (KNN), Random Forest (RF), and Support Vector Machines (SVM). The final accuracy of ResNet is 99.70%, which is higher than the 94.33% correct rate of the popular machine learning model SVM. Therefore, the electronic nose combined with conversion to images shows great potential for using deep transfer learning methods for chicken freshness classification. [ABSTRACT FROM AUTHOR]

Antimicrobial resistance (AMR) poses a threat to global public health. To mitigate the impacts of AMR, it is important to identify the molecular mechanisms of AMR and thereby determine optimal therapy as early as possible. Conventional machine learning-based drug-resistance analyses assume genetic variations to be homogeneous, thus not distinguishing between coding and intergenic sequences. In this study, we represent genetic data from Mycobacterium tuberculosis as a graph, and then adopt a deep graph learning method—heterogeneous graph attention network ('HGAT–AMR')—to predict anti-tuberculosis (TB) drug resistance. The HGAT–AMR model is able to accommodate incomplete phenotypic profiles, as well as provide 'attention scores' of genes and single nucleotide polymorphisms (SNPs) both at a population level and for individual samples. These scores encode the inputs, which the model is 'paying attention to' in making its drug resistance predictions. The results show that the proposed model generated the best area under the receiver operating characteristic (AUROC) for isoniazid and rifampicin (98.53 and 99.10%), the best sensitivity for three first-line drugs (94.91% for isoniazid, 96.60% for ethambutol and 90.63% for pyrazinamide), and maintained performance when the data were associated with incomplete phenotypes (i.e. for those isolates for which phenotypic data for some drugs were missing). We also demonstrate that the model successfully identifies genes and SNPs associated with drug resistance, mitigating the impact of resistance profile while considering particular drug resistance, which is consistent with domain knowledge. [ABSTRACT FROM AUTHOR]

The purpose of the study was to develop and clinically deploy an automated, deep learning‐based approach to treatment planning for whole‐brain radiotherapy (WBRT). We collected CT images and radiotherapy treatment plans to automate a beam aperture definition from 520 patients who received WBRT. These patients were split into training (n = 312), cross‐validation (n = 104), and test (n = 104) sets which were used to train and evaluate a deep learning model. The DeepLabV3+ architecture was trained to automatically define the beam apertures on lateral‐opposed fields using digitally reconstructed radiographs (DRRs). For the beam aperture evaluation, 1st quantitative analysis was completed using a test set before clinical deployment and 2nd quantitative analysis was conducted 90 days after clinical deployment. The mean surface distance and the Hausdorff distances were compared in the anterior‐inferior edge between the clinically used and the predicted fields. Clinically used plans and deep‐learning generated plans were evaluated by various dose–volume histogram metrics of brain, cribriform plate, and lens. The 1st quantitative analysis showed that the average mean surface distance and Hausdorff distance were 7.1 mm (±3.8 mm) and 11.2 mm (±5.2 mm), respectively, in the anterior–inferior edge of the field. The retrospective dosimetric comparison showed that brain dose coverage (D99%, D95%, D1%) of the automatically generated plans was 29.7, 30.3, and 32.5 Gy, respectively, and the average dose of both lenses was up to 19.0% lower when compared to the clinically used plans. Following the clinical deployment, the 2nd quantitative analysis showed that the average mean surface distance and Hausdorff distance between the predicted and clinically used fields were 2.6 mm (±3.2 mm) and 4.5 mm (±5.6 mm), respectively. In conclusion, the automated patient‐specific treatment planning solution for WBRT was implemented in our clinic. The predicted fields appeared consistent with clinically used fields and the predicted plans were dosimetrically comparable. [ABSTRACT FROM AUTHOR]

Purpose: The goal of this study was to predict soft tissue sarcoma response to radiotherapy (RT) using longitudinal diffusion‐weighted MRI (DWI). A novel deep‐learning prediction framework along with generative adversarial network (GAN)‐based data augmentation was investigated for the response prediction. Methods: Thirty soft tissue sarcoma patients who were treated with five‐fraction hypofractionated radiation therapy (RT, 6Gy×5) underwent diffusion‐weighted MRI three times throughout the RT course using an MR‐guided radiotherapy system. Pathologic treatment effect (TE) scores, ranging from 0‐100%, were obtained from the post‐RT surgical specimen as a surrogate of patient treatment response. Patients were divided into three classes based on the TE score (TE ≤ 20%, 20% < TE < 90%, TE ≥ 90%). Apparent diffusion coefficient (ADC) maps of the tumor from the three time points were combined as 3‐channel images. An auxiliary classifier generative adversarial network (ACGAN) was trained on 20 patients to augment the data size. A total of 15,000 synthetic images were generated for each class. A prediction model based on a previously described VGG‐19 network was trained using the synthesized data, validated on five unseen validation patients, and tested on the remaining five test patients. The entire process was repeated seven times, each time shuffling the training, validation, and testing datasets such that each patient was tested at least once during the independent test stage. Prediction performance for slice‐based prediction and patient‐based prediction was evaluated. Results: The average training and validation accuracies were 86.5% ± 1.6% and 84.8% ± 1.8%, respectively, indicating that the generated samples were good representations of the original patient data. Among the seven rounds of testing, slice by slice prediction accuracy ranged from 81.6% to 86.8%. The overall accuracy of the independent test sets was 83.3%. For patient‐based prediction, 80% was achieved in one round and 100% was achieved in the remaining six rounds. The mean accuracy was 97.1%. Conclusion: This study demonstrated the potential to use deep learning to predict the pathologic treatment effect from longitudinal DWI. Accuracies of 83.3% and 97.1% were achieved on independent test sets for slice‐based and patient‐based prediction respectively. [ABSTRACT FROM AUTHOR]

Edge caching is a promising method to deal with the traffic explosion problem towards future network. In order to satisfy the demands of user requests, the contents can be proactively cached locally at the proximity to users (e.g., base stations or user device). Recently, some learning-based edge caching optimizations are discussed. However, most of the previous studies explore the influence of dynamic and constant expanding action and caching space, leading to unpracticality and low efficiency. In this paper, we study the edge caching optimization problem by utilizing the Double Deep Q-network (Double DQN) learning framework to maximize the hit rate of user requests. Firstly, we obtain the Device-to-Device (D2D) sharing model by considering both online and offline factors and then we formulate the optimization problem, which is proved as NP-hard. Then the edge caching replacement problem is derived by Markov decision process (MDP). Finally, an edge caching strategy based on Double DQN is proposed. The experimental results based on large-scale actual traces show the effectiveness of the proposed framework. [ABSTRACT FROM AUTHOR]

Detection of fake news has spurred widespread interests in areas such as healthcare and Internet societies, in order to prevent propagating misleading information for commercial and political purposes. However, efforts to study a general framework for exploiting knowledge, for judging the trustworthiness of given news based on their content, have been limited. Indeed, the existing works rarely consider incorporating knowledge graphs (KGs), which could provide rich structured knowledge for better language understanding. In this work, we propose a deep triple network (DTN) that leverages knowledge graphs to facilitate fake news detection with triple-enhanced explanations. In the DTN, background knowledge graphs, such as open knowledge graphs and extracted graphs from news bases, are applied for both low-level and high-level feature extraction to classify the input news article and provide explanations for the classification. The performance of the proposed method is evaluated by demonstrating abundant convincing comparative experiments. Obtained results show that DTN outperforms conventional fake news detection methods from different aspects, including the provision of factual evidence supporting the decision of fake news detection. [ABSTRACT FROM AUTHOR]

The field of mechanical fault diagnosis has entered the era of "big data". However, existing diagnostic algorithms, relying on artificial feature extraction and expert knowledge are of poor extraction ability and lack self-adaptability in the mass data. In the fault diagnosis of rotating machinery, due to the accidental occurrence of equipment faults, the proportion of fault samples is small, the samples are imbalanced, and available data are scarce, which leads to the low accuracy rate of the intelligent diagnosis model trained to identify the equipment state. To solve the above problems, an end-to-end diagnosis model is first proposed, which is an intelligent fault diagnosis method based on one-dimensional convolutional neural network (1D-CNN). That is to say, the original vibration signal is directly input into the model for identification. After that, through combining the convolutional neural network with the generative adversarial networks, a data expansion method based on the one-dimensional deep convolutional generative adversarial networks (1D-DCGAN) is constructed to generate small sample size fault samples and construct the balanced data set. Meanwhile, in order to solve the problem that the network is difficult to optimize, gradient penalty and Wasserstein distance are introduced. Through the test of bearing database and hydraulic pump, it shows that the one-dimensional convolution operation has strong feature extraction ability for vibration signals. The proposed method is very accurate for fault diagnosis of the two kinds of equipment, and high-quality expansion of the original data can be achieved. [ABSTRACT FROM AUTHOR]

• A visual multimodal facial database for hedonic recognition is constructed. • Hedonic rating categories are identified by automatic facial expression recognition technique. • A hybrid deep learning framework is employed to exploit facial information. The emotional reactions that certain products evoke (like or dislike) play a key role in the user acceptance or preference toward these products and influence the choice behaviors. Facial expressions constitute a primary channel to express and communicate emotions. In particular, the facial reactions evoked by certain stimuli are generated subconsciously, and they can be easily captured by video recording. However, it is challenging to quantify and objectively describe facial expressions. Until now, most implicit measurements based on facial expression analyses have been implemented considering "basic emotions", which cannot directly represent like/dislike. How can we identify the affective state of consumers directly by using facial expression analysis technology? In this study, hedonic rating categories are defined based on the hedonic scale, and automatic facial expression analysis technology is applied to evaluate the suitability of the defined categories. Moreover, a visual multimodal facial database (RGB, depth, and infrared images) for the hedonic recognition of Asian individuals is constructed. A hybrid deep learning framework is employed to exploit the spatiotemporal information of the collected data to realize hedonic recognition. Early and late fusion strategies are used to investigate the suitability of the individual modalities and their complementarity. The experimental results indicate that the late fusion of the three modalities yielded the highest recognition accuracy compared to those obtained using individual modalities and the early fusion scheme. These results demonstrate the effectiveness of fusing different visual facial data and indicate that the use of multimodality-based facial expression analyses could enhance the accuracy of hedonic recognition. [ABSTRACT FROM AUTHOR]

There exist some problems of the traditional forest fire recognition technology, such as the complex background of the forest fire images, the weak generalization ability of the image recognition, and the low accuracy, which will lead to false alarm or missing alarm. To solve the problems mentioned above, a multi-level forest fire detection method based on depth learning has been proposed here. First, in order to solve the problem of uneven distribution and small number of samples, the high-quality forest fire samples have been generated with GAN (General Advanced Networks). Second, Adaboost classifier based on the characteristics of HOG (Histogram of Oriented Gridients) has been used to make primary prediction of forest fire area image, and then convolutional neural networks (CNN) and support vector machine (SVM) have been used to carry out the secondary recognition of the fire area. The experimental results show that the forest fire image recognition method proposed in this paper can obtain higher recognition rate and lower false alarm rate after training with fewer samples than other algorithms. At the same time, this method has lower requirements on the hardware environment required for the sample training and recognition, and has obvious advantages over algorithms that require GPU training environment. With this method, the recognition rate of forest fire images can reach 97.6 %, the false alarm rate is 1.4 %, and the missed alarm rate is 1%. The average time for recognizing sample pictures is only 0.7 s, which has high effectiveness and robustness. [ABSTRACT FROM AUTHOR]