Race against the Machine: can deep learning recognize microstructures as well as the trained human eye?

With the rapid development of technology, artificial intelligence is gradually applied in the field of image processing. Traditional image processing and recognition takes human extraction of features as a means to describe and characterize low-level edge information in images, but a large number of features extraction causes certain troubles to image recognition, and it is difficult to extract deep-level information. In recent years, deep learning, as an important emerging technology in the field of image recognition, has broad application prospects, and the research on image recognition technology has important theoretical value and practical significance to promote the development of computer vision and artificial intelligence. It opens up a whole new direction in image processing research, thus meeting the demand for efficient image processing. This paper reviews the applications of deep learning in image recognition, analyzes the classical convolutional neural network model, and summarizes the applications of deep learning in the fields of license plate recognition, face recognition, and medical image recognition.

Deep learning is a new field in machine learning, which is based on the neural network model to extract features by simulating human brain cognition. Image recognition is an essential subject in the field of deep learning and has made significant progress in recent years. In the training process, only the computer needs to cooperate, and no personnel need to participate in the process to get a good image recognition effect. This paper makes an in-depth analysis on the application of deep learning in image recognition, discusses the necessary steps of image trial and the basic principle of the neural network, and puts forward some problems and popular deep learning models used in image recognition.

With the rapid development of high-tech in China, the field of artificial intelligence has also been promoted. Image recognition is an important topic in the field of artificial intelligence, which mainly includes two modules: classification recognition and feature extraction; At the same time, as an important research direction of AI, deep learning has made rapid progress in recent years. It has been widely used in image recognition, speech recognition and other fields and has achieved great success. This paper analyzes the application of deep learning in image recognition, mainly from the aspects of face recognition, remote sensing image classification, etc. Its purpose is to provide help for relevant practitioners, so as to promote the development of image recognition in the tide of AI development.

This research aims to study the deep learning applications in image recognition and classification tasks. The advantages and limitations of deep learning are explored by analysing existing deep learning algorithms and their applications on tasks such as using images to do classification, image segmentation, and target detection. In the experimental part, this paper evaluates the performance of deep learning in image recognition by using classical neural network models, including training and testing models on a large size of image datasets. The results show that neural network has good classification and detection capabilities in image recognition tasks and also achieves good results for image segmentation tasks. However, the model training process of shallow neural network models is time-consuming and performs poorly for small-scale datasets. Based on these situations, this paper proposes some optimization strategies to achieve high performance and efficiency of neural networks in image recognition and analyze the performance differences between different strategies .

Deep learning has been the core topic in machine learning and convolutional neural network is one of the most prominent approaches. Convolutional neural network has won numerous competitions in recent years. It has outstanding results in image recognition. We review the different deep learning approaches which have been used in the field of image classification and localization.

This paper researches the localization method based on the finer-grained channel parameters using a deep learning model with a single access point. The basic principle is to transform the localization problem into a classification problem by using the advantages of deep learning in image recognition. First, we use the 2D recursively applied and projected-multiple signal classification algorithm to estimate the path parameters, and then the signal spectrum in 2D space is obtained which is plotted in a image. Then we present a structure based on convolutional neural network to obtain the local signal features and learn the correlation between its two-dimensional parameters and the coordinates of the target. Finally, we use the commercial-grade ray tracing software to evaluate our proposed methods, and the simulation results show that the proposed method can locate the target accurately using only a single access point.

Big data technology has not only become an important driving force of industry digitalization and intelligence, but also a fundamental source of innovation and development. Through the study of pricing model, the application of big data technology in service trading can be more standardized and orderly. As a special data form in the recommendation field, traditional recommendation methods are often difficult to model session sequence data. Although some scholars have improved the traditional recommendation method to solve the problem of session sequence modeling, the effectiveness of the recommendation system is limited by many problems in the model structure. In recent years, with deep learning in image recognition, natural language processing, speech recognition and other breakthroughs in the field of artificial intelligence, will study deep into the recommendation algorithm, can help to effectively solve the traditional recommendation algorithm of multi-source heterogeneous information, data sparse, cold start, data characteristics of the design of artificial rely on such issues, in this case, The results of traditional recommendation methods are often inaccurate, and the results are often lagging and repeatable.

Crater extraction and recognition is an important research content of deep space planetary science. Traditional crater detection algorithms (CDAs) are mainly based on crater feature construction which relies on high-quality data. With the application of deep learning in image recognition and semantic segmentation, new ideas have been brought to meteorite crater extraction. Many crater extraction algorithms based on artificial intelligence have been proposed which greatly simplifies the crater extraction process and improves detection accuracy. However, with the improvement of the accuracy, the convolution kernels become more and more, and the huge parameters and the consumption of storage and computing resources limit the application of these algorithms in mobile device. In order to solve this problem, we propose a compact crater extraction network based on model pruning. In the combination of U-Net and residual block, the network structure is optimized under the premise of retaining the longitude of large model extraction, and the balance of hardware resources and algorithm accuracy is achieved. The experimental results show that we compress the model to 4.9% of raw size and only lose almost 0.5% accuracy. It provides a reference for the application of CDAs on resource constrained platforms.

Simple SummaryMalignant melanomas (MMs) with aypical clinical presentation constitute a diagnostic pitfall, and false negatives carry the risk of a diagnostic delay and improper disease management. Among the most common, challenging presentation forms of MMs are those that clinically resemble seborrheic keratosis (SK). On the other hand, SK may mimic melanoma, producing ‘false positive overdiagnosis’ and leading to needless excisions. The evolving efficiency of deep learning algorithms in image recognition and the availability of large image databases have accelerated the development of advanced computer-aided systems for melanoma detection. In the present study, we used image data from the International Skin Image Collaboration archive to explore the capacity of deep knowledge transfer in the challenging diagnostic task of the atypical skin tumors of MM and SK.Malignant melanomas resembling seborrheic keratosis (SK-like MMs) are atypical, challenging to diagnose melanoma cases that carry the risk of delayed diagnosis and inadequate treatment. On the other hand, SK may mimic melanoma, producing a ‘false positive’ with unnecessary lesion excisions. The present study proposes a computer-based approach using dermoscopy images for the characterization of SΚ-like MMs. Dermoscopic images were retrieved from the International Skin Imaging Collaboration archive. Exploiting image embeddings from pretrained convolutional network VGG16, we trained a support vector machine (SVM) classification model on a data set of 667 images. SVM optimal hyperparameter selection was carried out using the Bayesian optimization method. The classifier was tested on an independent data set of 311 images with atypical appearance: MMs had an absence of pigmented network and had an existence of milia-like cysts. SK lacked milia-like cysts and had a pigmented network. Atypical MMs were characterized with a sensitivity and specificity of 78.6% and 84.5%, respectively. The advent of deep learning in image recognition has attracted the interest of computer science towards improved skin lesion diagnosis. Open-source, public access archives of skin images empower further the implementation and validation of computer-based systems that might contribute significantly to complex clinical diagnostic problems such as the characterization of SK-like MMs.

Deep learning (DL) model training must address the memory bottleneck to continue scaling. Processing-in-memory approaches can be a viable solution as they move computations near or into the memory, reducing substantial data movement. However, to deploy applications on such hardware, end-to-end software support is crucial for efficient computation mapping and scheduling as well as extensible code generation, but no consideration has been made for DL training workloads. In this paper, we propose XLA-NDP, a compiler and runtime solution for NDPX, a near-data processing (NDP) architecture integrated with an existing DL training framework. XLA-NDP offloads NDPX kernels and schedules them to overlap with GPU kernels to maximize parallelism based on GPU and NDPX costs, while providing a template-based code generator with low-level optimizations. The experiments showed that XLA-NDP provides up to 1.41x speedup (1.24x on average) over the GPU baseline for four DL model training.

Recently, deep learning has been exploited in the field of medical image analysis. However, the training of deep learning models with medical images is time-consuming since most medical image data are three-dimensional volumes or high-resolution two-dimensional images. Moreover, the optimization of numerous hyperparameters strongly affects the performance of deep learning. If a framework for training deep learning with hyperparameter optimization on a supercomputer system can be realized, it is expected to accelerate the training of deep learning with medical images. In this study, we described our novel environment for training deep learning with medical images on the supercomputer system in our institute (Reedbush-H supercomputer system) based on asynchronous parallel Bayesian optimization. We trained two types of automated lesion detection application in a constructed environment. The constructed environment enabled us to train deep learning with hyperparameter tuning in a short time.

Curious Containers: A framework for computational reproducibility in life sciences with support for Deep Learning applications

Robust perception systems allow farm robots to recognize weeds and vegetation, enabling the selective application of fertilizers and herbicides to mitigate the environmental impact of traditional agricultural practices. Today's perception systems typically rely on deep learning to interpret sensor data for tasks such as distinguishing soil, crops, and weeds. These approaches usually require substantial amounts of manually labeled training data, which is often time-consuming and requires domain expertise. This paper aims to reduce this limitation and propose an automated labeling pipeline for crop-weed semantic image segmentation in managed agricultural fields. It allows the training of deep learning models without or with only limited manual labeling of images. Our system uses RGB images recorded with unmanned aerial or ground robots operating in the field to produce semantic labels exploiting the field row structure for spatially consistent labeling. We use the rows previously detected to identify multiple crop rows, reducing labeling errors and improving consistency. We further reduce labeling errors by assigning an "unknown" class to challenging-to-segment vegetation. We use evidential deep learning because it provides predictions uncertainty estimates that we use to refine and improve our predictions. In this way, the evidential deep learning assigns high uncertainty to the weed class, as it is often less represented in the training data, allowing us to use the uncertainty to correct the semantic predictions. Experimental results suggest that our approach outperforms general-purpose labeling methods applied to crop fields by a large margin and domain-specific approaches on multiple fields and crop species. Using our generated labels to train deep learning models boosts our prediction performance on previously unseen fields with respect to unseen crop species, growth stages, or different lighting conditions. We obtain an IoU of 88.6% on crops, and 22.7% on weeds for a managed field of sugarbeets, where fully supervised methods have 83.4% on crops and 33.5% on weeds and other unsupervised domain-specific methods get 54.6% on crops and 11.2% on weeds. Finally, our method allows fine-tuning models trained in a fully supervised fashion to improve their performance in unseen field conditions up to +17.6% in mean IoU without additional manual labeling.

Deep learning has rapidly advanced in both performance and development, but the training of deep learning models is often time-consuming and resource intensive, and needing adjustment on a large number of parameters. This research looks at whether or not Bio inspired optimization (GA and PSO), can help deepen deep learning itself, so to speak. By providing these nature inspired methods, the magic of automated model setting, these methods can help you to automatically find the best model setting for you to speed up training, increase accuracy, and reduce overfitting. Through employing GA and PSO on tasks such as hyperparameter tuning and optimization of models are deep learning models made more adaptive, efficient and robust. The techniques are then compared with traditional optimization methods and their use in building better deep learning models with less manual effort is demonstrated.

As a basic task and key link of space situational awareness, space target recognition has become crucial in threat analysis, communication reconnaissance and electronic countermeasures. Using the fingerprint features carried by the electromagnetic signal to recognize is an effective method. Because traditional radiation source recognition technologies are difficult to obtain satisfactory expert features, automatic feature extraction methods based on deep learning have become popular. Although many deep learning schemes have been proposed, most of them are only used to solve the inter-class separable problem and ignore the intra-class compactness. In addition, the openness of the real space may invalidate the existing closed-set recognition methods. In order to solve the above problems, inspired by the application of prototype learning in image recognition, we propose a novel method for recognizing space radiation sources based on a multi-scale residual prototype learning network (MSRPLNet). The method can be used for both the closed- and open-set recognition of space radiation sources. Furthermore, we also design a joint decision algorithm for an open-set recognition task to identify unknown radiation sources. To verify the effectiveness and reliability of the proposed method, we built a set of satellite signal observation and receiving systems in a real external environment and collected eight Iridium signals. The experimental results show that the accuracy of our proposed method can reach 98.34% and 91.04% for the closed- and open-set recognition of eight Iridium targets, respectively. Compared to similar research works, our method has obvious advantages.