Bi-level optimization, especially the gradient-based category, has been widely used in the deep learning community including hyperparameter optimization and meta-knowledge extraction. Bi-level optimization embeds one problem within another and the gradient-based category solves the outer-level task by computing the hypergradient, which is much more efficient than classical methods such as the evolutionary algorithm. In this survey, we first give a formal definition of the gradient-based bi-level optimization. Next, we delineate criteria to determine if a research problem is apt for bi-level optimization and provide a practical guide on structuring such problems into a bi-level optimization framework, a feature particularly beneficial for those new to this domain. More specifically, there are two formulations: the single-task formulation to optimize hyperparameters such as regularization parameters and the distilled data, and the multi-task formulation to extract meta-knowledge such as the model initialization. With a bi-level formulation, we then discuss four bi-level optimization solvers to update the outer variable including explicit gradient update, proxy update, implicit function update, and closed-form update. Finally, we wrap up the survey by highlighting two prospective future directions: (1) Effective Data Optimization for Science examined through the lens of task formulation. (2) Accurate Explicit Proxy Update analyzed from an optimization standpoint.
Deep learning model inference is a key service in many businesses and scientific discovery processes. This paper introduces RIBBON, a novel deep learning inference serving system that meets two competing objectives: quality-of-service (QoS) target and cost-effectiveness. The key idea behind RIBBON is to intelligently employ a diverse set of cloud computing instances (heterogeneous instances) to meet the QoS target and maximize cost savings. RIBBON devises a Bayesian Optimization-driven strategy that helps users build the optimal set of heterogeneous instances for their model inference service needs on cloud computing platforms -- and, RIBBON demonstrates its superiority over existing approaches of inference serving systems using homogeneous instance pools. RIBBON saves up to 16% of the inference service cost for different learning models including emerging deep learning recommender system models and drug-discovery enabling models.
Deep Learning predictions with measurable confidence are increasingly desirable for real-world problems, especially in high-risk settings. The Conformal Prediction (CP) framework is a versatile solution that guarantees a maximum error rate given minimal constraints. In this paper, we propose a novel conformal loss function that approximates the traditionally two-step CP approach in a single step. By evaluating and penalising deviations from the stringent expected CP output distribution, a Deep Learning model may learn the direct relationship between the input data and the conformal p-values. We carry out a comprehensive empirical evaluation to show our novel loss function's competitiveness for seven binary and multi-class prediction tasks on five benchmark datasets. On the same datasets, our approach achieves significant training time reductions up to 86% compared to Aggregated Conformal Prediction (ACP), while maintaining comparable approximate validity and predictive efficiency.
Powder-based additive manufacturing has transformed the manufacturing industry over the last decade. In Laser Powder Bed Fusion, a specific part is built in an iterative manner in which two-dimensional cross-sections are formed on top of each other by melting and fusing the proper areas of the powder bed. In this process, the behavior of the melt pool and its thermal field has a very important role in predicting the quality of the manufactured part and its possible defects. However, the simulation of such a complex phenomenon is usually very time-consuming and requires huge computational resources. Flow-3D is one of the software packages capable of executing such simulations using iterative numerical solvers. In this work, we create three datasets of single-trail processes using Flow-3D and use them to train a convolutional neural network capable of predicting the behavior of the three-dimensional thermal field of the melt pool solely by taking three parameters as input: laser power, laser velocity, and time step. The CNN achieves a relative Root Mean Squared Error of 2% to 3% for the temperature field and an average Intersection over Union score of 80% to 90% in predicting the melt pool area. Moreover, since time is included as one of the inputs of the model, the thermal field can be instantly obtained for any arbitrary time step without the need to iterate and compute all the steps
Artificial Intelligence has enabled the implementation of more accurate and efficient solutions to problems in various areas. In the agricultural sector, one of the main needs is to know at all times the extent of land occupied or not by crops in order to improve production and profitability. The traditional methods of calculation demand the collection of data manually and in person in the field, causing high labor costs, execution times, and inaccuracy in the results. The present work proposes a new method based on Deep Learning techniques complemented with conventional programming for the determination of the area of populated and unpopulated crop areas. We have considered as a case study one of the most recognized companies in the planting and harvesting of sugar cane in Ecuador. The strategy combines a Generative Adversarial Neural Network (GAN) that is trained on a dataset of aerial photographs of natural and urban landscapes to improve image resolution; a Convolutional Neural Network (CNN) trained on a dataset of aerial photographs of sugar cane plots to distinguish populated or unpopulated crop areas; and a standard image processing module for the calculation of areas in a percentage manner. The experiments performed demonstrate a significant improvement in the quality of the aerial photographs as well as a remarkable differentiation between populated and unpopulated crop areas, consequently, a more accurate result of cultivated and uncultivated areas. The proposed method can be extended to the detection of possible pests, areas of weed vegetation, dynamic crop development, and both qualitative and quantitative quality control.
In this paper we propose a Long Short-Term Memory Network based method to forecast the energy consumption in public buildings, based on past measurements. Our approach consists of three main steps: data processing step, training and validation step, and finally the forecasting step. We tested our method on a data set consisting of measurements taken every half an hour from the main building of the National Archives of the United Kingdom, in Kew and as evaluation metrics we have used Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE).
In videos, the human's actions are of three-dimensional (3D) signals. These videos investigate the spatiotemporal knowledge of human behavior. The promising ability is investigated using 3D convolution neural networks (CNNs). The 3D CNNs have not yet achieved high output for their well-established two-dimensional (2D) equivalents in still photographs. Board 3D Convolutional Memory and Spatiotemporal fusion face training difficulty preventing 3D CNN from accomplishing remarkable evaluation. In this paper, we implement Hybrid Deep Learning Architecture that combines STIP and 3D CNN features to enhance the performance of 3D videos effectively. After implementation, the more detailed and deeper charting for training in each circle of space-time fusion. The training model further enhances the results after handling complicated evaluations of models. The video classification model is used in this implemented model. Intelligent 3D Network Protocol for Multimedia Data Classification using Deep Learning is introduced to further understand spacetime association in human endeavors. In the implementation of the result, the well-known dataset, i.e., UCF101 to, evaluates the performance of the proposed hybrid technique. The results beat the proposed hybrid technique that substantially beats the initial 3D CNNs. The results are compared with state-of-the-art frameworks from literature for action recognition on UCF101 with an accuracy of 95%.
This paper studies inference acceleration using distributed convolutional neural networks (CNNs) in collaborative edge computing. To ensure inference accuracy in inference task partitioning, we consider the receptive-field when performing segment-based partitioning. To maximize the parallelization between the communication and computing processes, thereby minimizing the total inference time of an inference task, we design a novel task collaboration scheme in which the overlapping zone of the sub-tasks on secondary edge servers (ESs) is executed on the host ES, named as HALP. We further extend HALP to the scenario of multiple tasks. Experimental results show that HALP can accelerate CNN inference in VGG-16 by 1.7-2.0x for a single task and 1.7-1.8x for 4 tasks per batch on GTX 1080TI and JETSON AGX Xavier, which outperforms the state-of-the-art work MoDNN. Moreover, we evaluate the service reliability under time-variant channel, which shows that HALP is an effective solution to ensure high service reliability with strict service deadline.
In recent years, Generative Adversarial Networks (GANs) have become a hot topic among researchers and engineers that work with deep learning. It has been a ground-breaking technique which can generate new pieces of content of data in a consistent way. The topic of GANs has exploded in popularity due to its applicability in fields like image generation and synthesis, and music production and composition. GANs have two competing neural networks: a generator and a discriminator. The generator is used to produce new samples or pieces of content, while the discriminator is used to recognize whether the piece of content is real or generated. What makes it different from other generative models is its ability to learn unlabeled samples. In this review paper, we will discuss the evolution of GANs, several improvements proposed by the authors and a brief comparison between the different models. Index Terms generative adversarial networks, unsupervised learning, deep learning.
Accurate breast cancer diagnosis through mammography has the potential to save millions of lives around the world. Deep learning (DL) methods have shown to be very effective for mass detection in mammograms. Additional improvements of current DL models will further improve the effectiveness of these methods. A critical issue in this context is how to pick the right hyperparameters for DL models. In this paper, we present GA-E2E, a new approach for tuning the hyperparameters of DL models for brest cancer detection using Genetic Algorithms (GAs). Our findings reveal that differences in parameter values can considerably alter the area under the curve (AUC), which is used to determine a classifier's performance.
Human fall is one of the very critical health issues, especially for elders and disabled people living alone. The number of elder populations is increasing steadily worldwide. Therefore, human fall detection is becoming an effective technique for assistive living for those people. For assistive living, deep learning and computer vision have been used largely. In this review article, we discuss deep learning (DL)-based state-of-the-art non-intrusive (vision-based) fall detection techniques. We also present a survey on fall detection benchmark datasets. For a clear understanding, we briefly discuss different metrics which are used to evaluate the performance of the fall detection systems. This article also gives a future direction on vision-based human fall detection techniques.
As electro-optical energy from the sun propagates through the atmosphere it is affected by radiative transfer effects including absorption, emission, and scattering. Modeling these affects is essential for scientific remote sensing measurements of the earth and atmosphere. For example, hyperspectral imagery is a form of digital imagery collected with many, often hundreds, of wavelengths of light in pixel. The amount of light measured at the sensor is the result of emitted sunlight, atmospheric radiative transfer, and the reflectance off the materials on the ground, all of which vary per wavelength resulting from multiple physical phenomena. Therefore measurements of the ground spectra or atmospheric constituents requires separating these different contributions per wavelength. In this paper, we create an autoencoder similar to denoising autoencoders treating the atmospheric affects as 'noise' and ground reflectance as truth per spectrum. We generate hundreds of thousands of training samples by taking random samples of spectra from laboratory measurements and adding atmospheric affects using physics-based modelling via MODTRAN (http://modtran.spectral.com/modtran\_home) by varying atmospheric inputs. This process ideally could create an autoencoder that would separate atmospheric effects and ground reflectance in hyperspectral imagery, a process called atmospheric compensation which is difficult and time-consuming requiring a combination of heuristic approximations, estimates of physical quantities, and physical modelling. While the accuracy of our method is not as good as other methods in the field, this an important first step in applying the growing field of deep learning of physical principles to atmospheric compensation in hyperspectral imagery and remote sensing.
Trajectory prediction has been widely pursued in many fields, and many model-based and model-free methods have been explored. The former include rule-based, geometric or optimization-based models, and the latter are mainly comprised of deep learning approaches. In this paper, we propose a new method combining both methodologies based on a new Neural Differential Equation model. Our new model (Neural Social Physics or NSP) is a deep neural network within which we use an explicit physics model with learnable parameters. The explicit physics model serves as a strong inductive bias in modeling pedestrian behaviors, while the rest of the network provides a strong data-fitting capability in terms of system parameter estimation and dynamics stochasticity modeling. We compare NSP with 15 recent deep learning methods on 6 datasets and improve the state-of-the-art performance by 5.56%-70%. Besides, we show that NSP has better generalizability in predicting plausible trajectories in drastically different scenarios where the density is 2-5 times as high as the testing data. Finally, we show that the physics model in NSP can provide plausible explanations for pedestrian behaviors, as opposed to black-box deep learning. Code is available: https://github.com/realcrane/Human-Trajectory-Prediction-via-Neural-Social-Physics.
Bone age assessment (BAA) is a standard method for determining the age difference between skeletal and chronological age. Manual processes are complicated and necessitate the expertise of experts. This is where deep learning comes into play. In this study, pre-trained models like VGG-16, InceptionV3, XceptionNet, and MobileNet are used to assess the bone age of the input data, and their mean average errors are compared and evaluated to see which model predicts the best.
MOTIVATION: Detection of prostate cancer during transrectal ultrasound-guided biopsy is challenging. The highly heterogeneous appearance of cancer, presence of ultrasound artefacts, and noise all contribute to these difficulties. Recent advancements in high-frequency ultrasound imaging - micro-ultrasound - have drastically increased the capability of tissue imaging at high resolution. Our aim is to investigate the development of a robust deep learning model specifically for micro-ultrasound-guided prostate cancer biopsy. For the model to be clinically adopted, a key challenge is to design a solution that can confidently identify the cancer, while learning from coarse histopathology measurements of biopsy samples that introduce weak labels. METHODS: We use a dataset of micro-ultrasound images acquired from 194 patients, who underwent prostate biopsy. We train a deep model using a co-teaching paradigm to handle noise in labels, together with an evidential deep learning method for uncertainty estimation. We evaluate the performance of our model using the clinically relevant metric of accuracy vs. confidence. RESULTS: Our model achieves a well-calibrated estimation of predictive uncertainty with area under the curve of 88$\%$. The use of co-teaching and evidential deep learning in combination yields significantly better uncertainty estimation than either alone. We also provide a detailed comparison against state-of-the-art in uncertainty estimation.
Automatic modulation recognition (AMR) detects the modulation scheme of the received signals for further signal processing without needing prior information, and provides the essential function when such information is missing. Recent breakthroughs in deep learning (DL) have laid the foundation for developing high-performance DL-AMR approaches for communications systems. Comparing with traditional modulation detection methods, DL-AMR approaches have achieved promising performance including high recognition accuracy and low false alarms due to the strong feature extraction and classification abilities of deep neural networks. Despite the promising potential, DL-AMR approaches also bring concerns to complexity and explainability, which affect the practical deployment in wireless communications systems. This paper aims to present a review of the current DL-AMR research, with a focus on appropriate DL models and benchmark datasets. We further provide comprehensive experiments to compare the state of the art models for single-input-single-output (SISO) systems from both accuracy and complexity perspectives, and propose to apply DL-AMR in the new multiple-input-multiple-output (MIMO) scenario with precoding. Finally, existing challenges and possible future research directions are discussed.
3D image segmentation is a recent and crucial step in many medical analysis and recognition schemes. In fact, it represents a relevant research subject and a fundamental challenge due to its importance and influence. This paper provides a multi-phase Deep Learning-based system that hybridizes various efficient methods in order to get the best 3D segmentation output. First, to reduce the amount of data and accelerate the processing time, the application of Decimate compression technique is suggested and justified. We then use a CNN model to segment dental images into fifteen separated classes. In the end, a special KNN-based transformation is applied for the purpose of removing isolated meshes and of correcting dental forms. Experimentations demonstrate the precision and the robustness of the selected framework applied to 3D dental images within a private clinical benchmark.
Designing and generating new data under targeted properties has been attracting various critical applications such as molecule design, image editing and speech synthesis. Traditional hand-crafted approaches heavily rely on expertise experience and intensive human efforts, yet still suffer from the insufficiency of scientific knowledge and low throughput to support effective and efficient data generation. Recently, the advancement of deep learning has created the opportunity for expressive methods to learn the underlying representation and properties of data. Such capability provides new ways of determining the mutual relationship between the structural patterns and functional properties of the data and leveraging such relationships to generate structural data, given the desired properties. This article is a systematic review that explains this promising research area, commonly known as controllable deep data generation. First, the article raises the potential challenges and provides preliminaries. Then the article formally defines controllable deep data generation, proposes a taxonomy on various techniques and summarizes the evaluation metrics in this specific domain. After that, the article introduces exciting applications of controllable deep data generation, experimentally analyzes and compares existing works. Finally, this article highlights the promising future directions of controllable deep data generation and identifies five potential challenges.
The ever-increasing computational complexity of deep learning models makes their training and deployment difficult on various cloud and edge platforms. Replacing floating-point arithmetic with low-bit integer arithmetic is a promising approach to save energy, memory footprint, and latency of deep learning models. As such, quantization has attracted the attention of researchers in recent years. However, using integer numbers to form a fully functional integer training pipeline including forward pass, back-propagation, and stochastic gradient descent is not studied in detail. Our empirical and mathematical results reveal that integer arithmetic seems to be enough to train deep learning models. Unlike recent proposals, instead of quantization, we directly switch the number representation of computations. Our novel training method forms a fully integer training pipeline that does not change the trajectory of the loss and accuracy compared to floating-point, nor does it need any special hyper-parameter tuning, distribution adjustment, or gradient clipping. Our experimental results show that our proposed method is effective in a wide variety of tasks such as classification (including vision transformers), object detection, and semantic segmentation.
Since the first success of Dong et al., the deep-learning-based approach has become dominant in the field of single-image super-resolution. This replaces all the handcrafted image processing steps of traditional sparse-coding-based methods with a deep neural network. In contrast to sparse-coding-based methods, which explicitly create high/low-resolution dictionaries, the dictionaries in deep-learning-based methods are implicitly acquired as a nonlinear combination of multiple convolutions. One disadvantage of deep-learning-based methods is that their performance is degraded for images created differently from the training dataset (out-of-domain images). We propose an end-to-end super-resolution network with a deep dictionary (SRDD), where a high-resolution dictionary is explicitly learned without sacrificing the advantages of deep learning. Extensive experiments show that explicit learning of high-resolution dictionary makes the network more robust for out-of-domain test images while maintaining the performance of the in-domain test images.