PyTorchKR Papers

최신 논문

총 8,922편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2111.14416

Being Patient and Persistent: Optimizing An Early Stopping Strategy for Deep Learning in Profiled Attacks

The absence of an algorithm that effectively monitors deep learning models used in side-channel attacks increases the difficulty of evaluation. If the attack is unsuccessful, the question is if we are dealing with a resistant implementation or a faulty model. We propose an early stopping algorithm that reliably recognizes the model's optimal state during training. The novelty of our solution is an efficient implementation of guessing entropy estimation. Additionally, we formalize two conditions, persistence and patience, for a deep learning model to be optimal. As a result, the model converges with fewer traces.

arXiv:2111.14247

A Survey of Large-Scale Deep Learning Serving System Optimization: Challenges and Opportunities

Deep Learning (DL) models have achieved superior performance in many application domains, including vision, language, medical, commercial ads, entertainment, etc. With the fast development, both DL applications and the underlying serving hardware have demonstrated strong scaling trends, i.e., Model Scaling and Compute Scaling, for example, the recent pre-trained model with hundreds of billions of parameters with ~TB level memory consumption, as well as the newest GPU accelerators providing hundreds of TFLOPS. With both scaling trends, new problems and challenges emerge in DL inference serving systems, which gradually trends towards Large-scale Deep learning Serving systems (LDS). This survey aims to summarize and categorize the emerging challenges and optimization opportunities for large-scale deep learning serving systems. By providing a novel taxonomy, summarizing the computing paradigms, and elaborating the recent technique advances, we hope that this survey could shed light on new optimization perspectives and motivate novel works in large-scale deep learning system optimization.

arXiv:2111.14051

Enabling Fast Deep Learning on Tiny Energy-Harvesting IoT Devices

Energy harvesting (EH) IoT devices that operate intermittently without batteries, coupled with advances in deep neural networks (DNNs), have opened up new opportunities for enabling sustainable smart applications. Nevertheless, implementing those computation and memory-intensive intelligent algorithms on EH devices is extremely difficult due to the challenges of limited resources and intermittent power supply that causes frequent failures. To address those challenges, this paper proposes a methodology that enables fast deep learning with low-energy accelerators for tiny energy harvesting devices. We first propose $RAD$, a resource-aware structured DNN training framework, which employs block circulant matrix and structured pruning to achieve high compression for leveraging the advantage of various vector operation accelerators. A DNN implementation method, $ACE$, is then proposed that employs low-energy accelerators to profit maximum performance with small energy consumption. Finally, we further design $FLEX$, the system support for intermittent computation in energy harvesting situations. Experimental results from three different DNN models demonstrate that $RAD$, $ACE$, and $FLEX$ can enable fast and correct inference on energy harvesting devices with up to 4.26X runtime reduction, up to 7.7X energy reduction with higher accuracy over the state-of-the-art.

arXiv:2111.13861

A New Multifractal-based Deep Learning Model for Text Mining

In this world full of uncertainty, where the fabric of existence weaves patterns of complexity, multifractal emerges as beacons of insight, illuminating them. As we delve into the realm of text mining that underpins various natural language processing applications and powers a range of intelligent services, we recognize that behind the veil of text lies a manifestation of human thought and cognition, intricately intertwined with the complexities. Building upon the foundation of perceiving text as a complex system, this study embarks on a journey to unravel the hidden treasures within, armed with the proposed multifractal method that deciphers the multifractal attributes embedded within the text landscape. This endeavor culminates in the birth of our novel model, which also harnesses the power of the proposed activation function to facilitate nonlinear information transmission within its neural network architecture. The success on experiments anchored in real-world technical reports covering the extraction of technical term and classification of hazard events, stands as a testament to our endeavors. This research venture not only expands our understanding of text mining but also opens new horizons for knowledge discovery across various domains.

arXiv:2111.14396

Lightweight Deep Learning Architecture for MPI Correction and Transient Reconstruction

Indirect Time-of-Flight cameras (iToF) are low-cost devices that provide depth images at an interactive frame rate. However, they are affected by different error sources, with the spotlight taken by Multi-Path Interference (MPI), a key challenge for this technology. Common data-driven approaches tend to focus on a direct estimation of the output depth values, ignoring the underlying transient propagation of the light in the scene. In this work instead, we propose a very compact architecture, leveraging on the direct-global subdivision of transient information for the removal of MPI and for the reconstruction of the transient information itself. The proposed model reaches state-of-the-art MPI correction performances both on synthetic and real data and proves to be very competitive also at extreme levels of noise; at the same time, it also makes a step towards reconstructing transient information from multi-frequency iToF data.

arXiv:2111.13785공식 코드 · Papers with Code

OmiTrans: generative adversarial networks based omics-to-omics translation framework

With the rapid development of high-throughput experimental technologies, different types of omics (e.g., genomics, epigenomics, transcriptomics, proteomics, and metabolomics) data can be produced from clinical samples. The correlations between different omics types attracts a lot of research interest, whereas the stduy on genome-wide omcis data translation (i.e, generation and prediction of one type of omics data from another type of omics data) is almost blank. Generative adversarial networks and the variants are one of the most state-of-the-art deep learning technologies, which have shown great success in image-to-image translation, text-to-image translation, etc. Here we proposed OmiTrans, a deep learning framework adopted the idea of generative adversarial networks to achieve omics-to-omics translation with promising results. OmiTrans was able to faithfully reconstruct gene expression profiles from DNA methylation data with high accuracy and great model generalisation, as demonstrated in the experiments.

arXiv:2111.14338공식 코드 · Papers with Code

Improving Deep Learning Interpretability by Saliency Guided Training

Saliency methods have been widely used to highlight important input features in model predictions. Most existing methods use backpropagation on a modified gradient function to generate saliency maps. Thus, noisy gradients can result in unfaithful feature attributions. In this paper, we tackle this issue and introduce a {\it saliency guided training}procedure for neural networks to reduce noisy gradients used in predictions while retaining the predictive performance of the model. Our saliency guided training procedure iteratively masks features with small and potentially noisy gradients while maximizing the similarity of model outputs for both masked and unmasked inputs. We apply the saliency guided training procedure to various synthetic and real data sets from computer vision, natural language processing, and time series across diverse neural architectures, including Recurrent Neural Networks, Convolutional Networks, and Transformers. Through qualitative and quantitative evaluations, we show that saliency guided training procedure significantly improves model interpretability across various domains while preserving its predictive performance.

arXiv:2111.14088

Multicriteria interpretability driven Deep Learning

Deep Learning methods are renowned for their performances, yet their lack of interpretability prevents them from high-stakes contexts. Recent model agnostic methods address this problem by providing post-hoc interpretability methods by reverse-engineering the model's inner workings. However, in many regulated fields, interpretability should be kept in mind from the start, which means that post-hoc methods are valid only as a sanity check after model training. Interpretability from the start, in an abstract setting, means posing a set of soft constraints on the model's behavior by injecting knowledge and annihilating possible biases. We propose a Multicriteria technique that allows to control the feature effects on the model's outcome by injecting knowledge in the objective function. We then extend the technique by including a non-linear knowledge function to account for more complex effects and local lack of knowledge. The result is a Deep Learning model that embodies interpretability from the start and aligns with the recent regulations. A practical empirical example based on credit risk, suggests that our approach creates performant yet robust models capable of overcoming biases derived from data scarcity.

arXiv:2111.13834

Federated Deep Learning in Electricity Forecasting: An MCDM Approach

Large-scale data analysis is growing at an exponential rate as data proliferates in our societies. This abundance of data has the advantage of allowing the decision-maker to implement complex models in scenarios that were prohibitive before. At the same time, such an amount of data requires a distributed thinking approach. In fact, Deep Learning models require plenty of resources, and distributed training is needed. This paper presents a Multicriteria approach for distributed learning. Our approach uses the Weighted Goal Programming approach in its Chebyshev formulation to build an ensemble of decision rules that optimize aprioristically defined performance metrics. Such a formulation is beneficial because it is both model and metric agnostic and provides an interpretable output for the decision-maker. We test our approach by showing a practical application in electricity demand forecasting. Our results suggest that when we allow for dataset split overlapping, the performances of our methodology are consistently above the baseline model trained on the whole dataset.

arXiv:2111.13813공식 코드 · Papers with Code

Video Content Classification using Deep Learning

Video content classification is an important research content in computer vision, which is widely used in many fields, such as image and video retrieval, computer vision. This paper presents a model that is a combination of Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) which develops, trains, and optimizes a deep learning network that can identify the type of video content and classify them into categories such as "Animation, Gaming, natural content, flat content, etc". To enhance the performance of the model novel keyframe extraction method is included to classify only the keyframes, thereby reducing the overall processing time without sacrificing any significant performance.

arXiv:2111.13519공식 코드 · Papers with Code

Graph Auto-Encoders for Financial Clustering

Deep learning has shown remarkable results on Euclidean data (e.g. audio, images, text) however this type of data is limited in the amount of relational information it can hold. In mathematics we can model more general relational data in a graph structure while retaining Euclidean data as associated node or edge features. Due to the ubiquity of graph data, and its ability to hold multiple dimensions of information, graph deep learning has become a fast emerging field. We look at applying and optimising graph deep learning on a finance graph to produce more informed clusters of companies. Having clusters produced from multiple streams of data can be highly useful in quantitative finance; not only does it allow clusters to be tailored to the specific task but the culmination of multiple streams allows for cross source pattern recognition that would have otherwise gone unnoticed. This can provide financial institutions with an edge over competitors which is crucial in the heavily optimised world of trading. In this paper we use news co-occurrence and stock price for our data combination. We optimise our model to achieve an average testing precision of 78% and find a clear improvement in clustering capabilities when dual data sources are used; cluster purity rises from 32% for just vertex data and 42% for just edge data to 64% when both are used in comparisons to ground-truth Bloomberg clusters. The framework we provide utilises unsupervised learning which we view as key for future work due to the volume of unlabelled data in financial markets.

arXiv:2111.13034

DeepWiVe: Deep-Learning-Aided Wireless Video Transmission

We present DeepWiVe, the first-ever end-to-end joint source-channel coding (JSCC) video transmission scheme that leverages the power of deep neural networks (DNNs) to directly map video signals to channel symbols, combining video compression, channel coding, and modulation steps into a single neural transform. Our DNN decoder predicts residuals without distortion feedback, which improves video quality by accounting for occlusion/disocclusion and camera movements. We simultaneously train different bandwidth allocation networks for the frames to allow variable bandwidth transmission. Then, we train a bandwidth allocation network using reinforcement learning (RL) that optimizes the allocation of limited available channel bandwidth among video frames to maximize overall visual quality. Our results show that DeepWiVe can overcome the cliff-effect, which is prevalent in conventional separation-based digital communication schemes, and achieve graceful degradation with the mismatch between the estimated and actual channel qualities. DeepWiVe outperforms H.264 video compression followed by low-density parity check (LDPC) codes in all channel conditions by up to 0.0462 on average in terms of the multi-scale structural similarity index measure (MS-SSIM), while beating H.265 + LDPC by up to 0.0058 on average. We also illustrate the importance of optimizing bandwidth allocation in JSCC video transmission by showing that our optimal bandwidth allocation policy is superior to the na\"ive uniform allocation. We believe this is an important step towards fulfilling the potential of an end-to-end optimized JSCC wireless video transmission system that is superior to the current separation-based designs.

arXiv:2111.12963

Error Bounds for a Matrix-Vector Product Approximation with Deep ReLU Neural Networks

Among the several paradigms of artificial intelligence (AI) or machine learning (ML), a remarkably successful paradigm is deep learning. Deep learning's phenomenal success has been hoped to be interpreted via fundamental research on the theory of deep learning. Accordingly, applied research on deep learning has spurred the theory of deep learning-oriented depth and breadth of developments. Inspired by such developments, we pose these fundamental questions: can we accurately approximate an arbitrary matrix-vector product using deep rectified linear unit (ReLU) feedforward neural networks (FNNs)? If so, can we bound the resulting approximation error? In light of these questions, we derive error bounds in Lebesgue and Sobolev norms that comprise our developed deep approximation theory. Guided by this theory, we have successfully trained deep ReLU FNNs whose test results justify our developed theory. The developed theory is also applicable for guiding and easing the training of teacher deep ReLU FNNs in view of the emerging teacher-student AI or ML paradigms that are essential for solving several AI or ML problems in wireless communications and signal processing; network science and graph signal processing; and network neuroscience and brain physics.

arXiv:2111.13663공식 코드 · Papers with Code

3D shape sensing and deep learning-based segmentation of strawberries

Automation and robotisation of the agricultural sector are seen as a viable solution to socio-economic challenges faced by this industry. This technology often relies on intelligent perception systems providing information about crops, plants and the entire environment. The challenges faced by traditional 2D vision systems can be addressed by modern 3D vision systems which enable straightforward localisation of objects, size and shape estimation, or handling of occlusions. So far, the use of 3D sensing was mainly limited to indoor or structured environments. In this paper, we evaluate modern sensing technologies including stereo and time-of-flight cameras for 3D perception of shape in agriculture and study their usability for segmenting out soft fruit from background based on their shape. To that end, we propose a novel 3D deep neural network which exploits the organised nature of information originating from the camera-based 3D sensors. We demonstrate the superior performance and efficiency of the proposed architecture compared to the state-of-the-art 3D networks. Through a simulated study, we also show the potential of the 3D sensing paradigm for object segmentation in agriculture and provide insights and analysis of what shape quality is needed and expected for further analysis of crops. The results of this work should encourage researchers and companies to develop more accurate and robust 3D sensing technologies to assure their wider adoption in practical agricultural applications.

arXiv:2111.12621

Accelerating Deep Learning with Dynamic Data Pruning

Deep learning's success has been attributed to the training of large, overparameterized models on massive amounts of data. As this trend continues, model training has become prohibitively costly, requiring access to powerful computing systems to train state-of-the-art networks. A large body of research has been devoted to addressing the cost per iteration of training through various model compression techniques like pruning and quantization. Less effort has been spent targeting the number of iterations. Previous work, such as forget scores and GraNd/EL2N scores, address this problem by identifying important samples within a full dataset and pruning the remaining samples, thereby reducing the iterations per epoch. Though these methods decrease the training time, they use expensive static scoring algorithms prior to training. When accounting for the scoring mechanism, the total run time is often increased. In this work, we address this shortcoming with dynamic data pruning algorithms. Surprisingly, we find that uniform random dynamic pruning can outperform the prior work at aggressive pruning rates. We attribute this to the existence of "sometimes" samples -- points that are important to the learned decision boundary only some of the training time. To better exploit the subtlety of sometimes samples, we propose two algorithms, based on reinforcement learning techniques, to dynamically prune samples and achieve even higher accuracy than the random dynamic method. We test all our methods against a full-dataset baseline and the prior work on CIFAR-10 and CIFAR-100, and we can reduce the training time by up to 2x without significant performance loss. Our results suggest that data pruning should be understood as a dynamic process that is closely tied to a model's training trajectory, instead of a static step based solely on the dataset alone.

arXiv:2111.12295

Animal behavior classification via deep learning on embedded systems

We develop an end-to-end deep-neural-network-based algorithm for classifying animal behavior using accelerometry data on the embedded system of an artificial intelligence of things (AIoT) device installed in a wearable collar tag. The proposed algorithm jointly performs feature extraction and classification utilizing a set of infinite-impulse-response (IIR) and finite-impulse-response (FIR) filters together with a multilayer perceptron. The utilized IIR and FIR filters can be viewed as specific types of recurrent and convolutional neural network layers, respectively. We evaluate the performance of the proposed algorithm via two real-world datasets collected from total eighteen grazing beef cattle using collar tags. The results show that the proposed algorithm offers good intra- and inter-dataset classification accuracy and outperforms its closest contenders including two state-of-the-art convolutional-neural-network-based time-series classification algorithms, which are significantly more complex. We implement the proposed algorithm on the embedded system of the utilized collar tags' AIoT device to perform in-situ classification of animal behavior. We achieve real-time in-situ behavior inference from accelerometry data without imposing any strain on the available computational, memory, or energy resources of the embedded system.

arXiv:2111.12123

MICS : Multi-steps, Inverse Consistency and Symmetric deep learning registration network

Deformable registration consists of finding the best dense correspondence between two different images. Many algorithms have been published, but the clinical application was made difficult by the high calculation time needed to solve the optimisation problem. Deep learning overtook this limitation by taking advantage of GPU calculation and the learning process. However, many deep learning methods do not take into account desirable properties respected by classical algorithms. In this paper, we present MICS, a novel deep learning algorithm for medical imaging registration. As registration is an ill-posed problem, we focused our algorithm on the respect of different properties: inverse consistency, symmetry and orientation conservation. We also combined our algorithm with a multi-step strategy to refine and improve the deformation grid. While many approaches applied registration to brain MRI, we explored a more challenging body localisation: abdominal CT. Finally, we evaluated our method on a dataset used during the Learn2Reg challenge, allowing a fair comparison with published methods.

논문 ID: 3713.31969

Deep Learning for Entity Matching: A Design Space Exploration

Entity matching (EM) finds data instances that refer to the same real-world entity. In this paper we examine applying deep learning (DL) to EM, to understand DL's benefits and limitations. We review many DL solutions that have been developed for related matching tasks in text processing (e.g., entity linking, textual entailment, etc.). We categorize these solutions and define a space of DL solutions for EM, as embodied by four solutions with varying representational power: SIF, RNN, Attention, and Hybrid. Next, we investigate the types of EM problems for which DL can be helpful. We consider three such problem types, which match structured data instances, textual instances, and dirty instances, respectively. We empirically compare the above four DL solutions with Magellan, a state-of-the-art learning-based EM solution. The results show that DL does not outperform current solutions on structured EM, but it can significantly outperform them on textual and dirty EM. For practitioners, this suggests that they should seriously consider using DL for textual and dirty EM problems. Finally, we analyze DL's performance and discuss future research directions.

arXiv:2111.11873

Deformable image registration with deep network priors: a study on longitudinal PET images

Longitudinal image registration is challenging and has not yet benefited from major performance improvements thanks to deep-learning. Inspired by Deep Image Prior, this paper introduces a different use of deep architectures as regularizers to tackle the image registration question. We propose a subject-specific deformable registration method called MIRRBA, relying on a deep pyramidal architecture to be the prior parametric model constraining the deformation field. Diverging from the supervised learning paradigm, MIRRBA does not require a learning database, but only the pair of images to be registered to optimize the network's parameters and provide a deformation field. We demonstrate the regularizing power of deep architectures and present new elements to understand the role of the architecture in deep learning methods for registration. Hence, to study the impact of the network parameters, we ran our method with different architectural configurations on a private dataset of 110 metastatic breast cancer full-body PET images with manual segmentations of the brain, bladder and metastatic lesions. We compared it against conventional iterative registration approaches and supervised deep learning-based models. Global and local registration accuracies were evaluated using the detection rate and the Dice score respectively, while registration realism was evaluated using the Jacobian's determinant. Moreover, we computed the ability of the different methods to shrink vanishing lesions with the disappearing rate. MIRRBA significantly improves the organ and lesion Dice scores of supervised models. Regarding the disappearing rate, MIRRBA more than doubles the best performing conventional approach SyNCC score. Our work therefore proposes an alternative way to bridge the performance gap between conventional and deep learning-based methods and demonstrates the regularizing power of deep architectures.

arXiv:2111.11828

Variance Reduction in Deep Learning: More Momentum is All You Need

Variance reduction (VR) techniques have contributed significantly to accelerating learning with massive datasets in the smooth and strongly convex setting (Schmidt et al., 2017; Johnson & Zhang, 2013; Roux et al., 2012). However, such techniques have not yet met the same success in the realm of large-scale deep learning due to various factors such as the use of data augmentation or regularization methods like dropout (Defazio & Bottou, 2019). This challenge has recently motivated the design of novel variance reduction techniques tailored explicitly for deep learning (Arnold et al., 2019; Ma & Yarats, 2018). This work is an additional step in this direction. In particular, we exploit the ubiquitous clustering structure of rich datasets used in deep learning to design a family of scalable variance reduced optimization procedures by combining existing optimizers (e.g., SGD+Momentum, Quasi Hyperbolic Momentum, Implicit Gradient Transport) with a multi-momentum strategy (Yuan et al., 2019). Our proposal leads to faster convergence than vanilla methods on standard benchmark datasets (e.g., CIFAR and ImageNet). It is robust to label noise and amenable to distributed optimization. We provide a parallel implementation in JAX.