PyTorchKR Papers

최신 논문

총 2,063편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2112.09245공식 코드 · Papers with Code

Automated Deep Learning: Neural Architecture Search Is Not the End

Deep learning (DL) has proven to be a highly effective approach for developing models in diverse contexts, including visual perception, speech recognition, and machine translation. However, the end-to-end process for applying DL is not trivial. It requires grappling with problem formulation and context understanding, data engineering, model development, deployment, continuous monitoring and maintenance, and so on. Moreover, each of these steps typically relies heavily on humans, in terms of both knowledge and interactions, which impedes the further advancement and democratization of DL. Consequently, in response to these issues, a new field has emerged over the last few years: automated deep learning (AutoDL). This endeavor seeks to minimize the need for human involvement and is best known for its achievements in neural architecture search (NAS), a topic that has been the focus of several surveys. That stated, NAS is not the be-all and end-all of AutoDL. Accordingly, this review adopts an overarching perspective, examining research efforts into automation across the entirety of an archetypal DL workflow. In so doing, this work also proposes a comprehensive set of ten criteria by which to assess existing work in both individual publications and broader research areas. These criteria are: novelty, solution quality, efficiency, stability, interpretability, reproducibility, engineering quality, scalability, generalizability, and eco-friendliness. Thus, ultimately, this review provides an evaluative overview of AutoDL in the early 2020s, identifying where future opportunities for progress may exist.

arXiv:2201.08295공식 코드 · Papers with Code

DIVA-DAF: A Deep Learning Framework for Historical Document Image Analysis

Deep learning methods have shown strong performance in solving tasks for historical document image analysis. However, despite current libraries and frameworks, programming an experiment or a set of experiments and executing them can be time-consuming. This is why we propose an open-source deep learning framework, DIVA-DAF, which is based on PyTorch Lightning and specifically designed for historical document analysis. Pre-implemented tasks such as segmentation and classification can be easily used or customized. It is also easy to create one's own tasks with the benefit of powerful modules for loading data, even large data sets, and different forms of ground truth. The applications conducted have demonstrated time savings for the programming of a document analysis task, as well as for different scenarios such as pre-training or changing the architecture. Thanks to its data module, the framework also allows to reduce the time of model training significantly.

arXiv:2201.07598공식 코드 · Papers with Code

Near-Optimal Sparse Allreduce for Distributed Deep Learning

Communication overhead is one of the major obstacles to train large deep learning models at scale. Gradient sparsification is a promising technique to reduce the communication volume. However, it is very challenging to obtain real performance improvement because of (1) the difficulty of achieving an scalable and efficient sparse allreduce algorithm and (2) the sparsification overhead. This paper proposes O$k$-Top$k$, a scheme for distributed training with sparse gradients. O$k$-Top$k$ integrates a novel sparse allreduce algorithm (less than 6$k$ communication volume which is asymptotically optimal) with the decentralized parallel Stochastic Gradient Descent (SGD) optimizer, and its convergence is proved. To reduce the sparsification overhead, O$k$-Top$k$ efficiently selects the top-$k$ gradient values according to an estimated threshold. Evaluations are conducted on the Piz Daint supercomputer with neural network models from different deep learning domains. Empirical results show that O$k$-Top$k$ achieves similar model accuracy to dense allreduce. Compared with the optimized dense and the state-of-the-art sparse allreduces, O$k$-Top$k$ is more scalable and significantly improves training throughput (e.g., 3.29x-12.95x improvement for BERT on 256 GPUs).

arXiv:2201.05867공식 코드 · Papers with Code

Transferability in Deep Learning: A Survey

The success of deep learning algorithms generally depends on large-scale data, while humans appear to have inherent ability of knowledge transfer, by recognizing and applying relevant knowledge from previous learning experiences when encountering and solving unseen tasks. Such an ability to acquire and reuse knowledge is known as transferability in deep learning. It has formed the long-term quest towards making deep learning as data-efficient as human learning, and has been motivating fruitful design of more powerful deep learning algorithms. We present this survey to connect different isolated areas in deep learning with their relation to transferability, and to provide a unified and complete view to investigating transferability through the whole lifecycle of deep learning. The survey elaborates the fundamental goals and challenges in parallel with the core principles and methods, covering recent cornerstones in deep architectures, pre-training, task adaptation and domain adaptation. This highlights unanswered questions on the appropriate objectives for learning transferable knowledge and for adapting the knowledge to new tasks and domains, avoiding catastrophic forgetting and negative transfer. Finally, we implement a benchmark and an open-source library, enabling a fair evaluation of deep learning methods in terms of transferability.

arXiv:2201.03767공식 코드 · Papers with Code

Improved (Related-key) Differential-based Neural Distinguishers for SIMON and SIMECK Block Ciphers

In CRYPTO 2019, Gohr made a pioneering attempt and successfully applied deep learning to the differential cryptanalysis against NSA block cipher SPECK32/64, achieving higher accuracy than the pure differential distinguishers. By its very nature, mining effective features in data plays a crucial role in data-driven deep learning. In this paper, in addition to considering the integrity of the information from the training data of the ciphertext pair, domain knowledge about the structure of differential cryptanalysis is also considered into the training process of deep learning to improve the performance. Meanwhile, taking the performance of the differential-neural distinguisher of SIMON32/64 as an entry point, we investigate the impact of input difference on the performance of the hybrid distinguishers to choose the proper input difference. Eventually, we improve the accuracy of the neural distinguishers of SIMON32/64, SIMON64/128, SIMECK32/64, and SIMECK64/128. We also obtain related-key differential-based neural distinguishers on round-reduced versions of SIMON32/64, SIMON64/128, SIMECK32/64, and SIMECK64/128 for the first time.

arXiv:2201.03115공식 코드 · Papers with Code

Semantic and sentiment analysis of selected Bhagavad Gita translations using BERT-based language framework

It is well known that translations of songs and poems not only break rhythm and rhyming patterns, but can also result in loss of semantic information. The Bhagavad Gita is an ancient Hindu philosophical text originally written in Sanskrit that features a conversation between Lord Krishna and Arjuna prior to the Mahabharata war. The Bhagavad Gita is also one of the key sacred texts in Hinduism and is known as the forefront of the Vedic corpus of Hinduism. In the last two centuries, there has been a lot of interest in Hindu philosophy from western scholars; hence, the Bhagavad Gita has been translated in a number of languages. However, there is not much work that validates the quality of the English translations. Recent progress of language models powered by deep learning has enabled not only translations but a better understanding of language and texts with semantic and sentiment analysis. Our work is motivated by the recent progress of language models powered by deep learning methods. In this paper, we present a framework that compares selected translations (from Sanskrit to English) of the Bhagavad Gita using semantic and sentiment analyses. We use hand-labelled sentiment dataset for tuning state-of-art deep learning-based language model known as bidirectional encoder representations from transformers (BERT). We provide sentiment and semantic analysis for selected chapters and verses across translations. Our results show that although the style and vocabulary in the respective translations vary widely, the sentiment analysis and semantic similarity shows that the message conveyed are mostly similar.

arXiv:2201.01871공식 코드 · Papers with Code

Direct multi-modal inversion of geophysical logs using deep learning

Geosteering of wells requires fast interpretation of geophysical logs, which is a non-unique inverse problem. Current work presents a proof-of-concept approach to multi-modal probabilistic inversion of logs using a single evaluation of an artificial deep neural network (DNN). A mixture density DNN (MDN) is trained using the "multiple-trajectory-prediction" (MTP) loss functions, which avoids mode collapse typical for traditional MDNs, and allows multi-modal prediction ahead of data. The proposed approach is verified on the real-time stratigraphic inversion of gamma-ray logs. The multi-modal predictor outputs several likely inverse solutions/predictions, providing more accurate and realistic solutions than a deterministic regression using a DNN. For these likely stratigraphic curves, the model simultaneously predicts their probabilities, which are implicitly learned from the training geological data. The stratigraphy predictions and their probabilities obtained in milliseconds from the MDN can enable better real-time decisions under geological uncertainties.

arXiv:2201.01203공식 코드 · Papers with Code

Quantifying Uncertainty in Deep Learning Approaches to Radio Galaxy Classification

In this work we use variational inference to quantify the degree of uncertainty in deep learning model predictions of radio galaxy classification. We show that the level of model posterior variance for individual test samples is correlated with human uncertainty when labelling radio galaxies. We explore the model performance and uncertainty calibration for different weight priors and suggest that a sparse prior produces more well-calibrated uncertainty estimates. Using the posterior distributions for individual weights, we demonstrate that we can prune 30% of the fully-connected layer weights without significant loss of performance by removing the weights with the lowest signal-to-noise ratio. A larger degree of pruning can be achieved using a Fisher information based ranking, but both pruning methods affect the uncertainty calibration for Fanaroff-Riley type I and type II radio galaxies differently. Like other work in this field, we experience a cold posterior effect, whereby the posterior must be down-weighted to achieve good predictive performance. We examine whether adapting the cost function to accommodate model misspecification can compensate for this effect, but find that it does not make a significant difference. We also examine the effect of principled data augmentation and find that this improves upon the baseline but also does not compensate for the observed effect. We interpret this as the cold posterior effect being due to the overly effective curation of our training sample leading to likelihood misspecification, and raise this as a potential issue for Bayesian deep learning approaches to radio galaxy classification in future.

arXiv:2112.09172공식 코드 · Papers with Code

An Audio-Visual Dataset and Deep Learning Frameworks for Crowded Scene Classification

This paper presents a task of audio-visual scene classification (SC) where input videos are classified into one of five real-life crowded scenes: 'Riot', 'Noise-Street', 'Firework-Event', 'Music-Event', and 'Sport-Atmosphere'. To this end, we firstly collect an audio-visual dataset (videos) of these five crowded contexts from Youtube (in-the-wild scenes). Then, a wide range of deep learning frameworks are proposed to deploy either audio or visual input data independently. Finally, results obtained from high-performed deep learning frameworks are fused to achieve the best accuracy score. Our experimental results indicate that audio and visual input factors independently contribute to the SC task's performance. Significantly, an ensemble of deep learning frameworks exploring either audio or visual input data can achieve the best accuracy of 95.7%.

arXiv:2112.11317공식 코드 · Papers with Code

Deep Learning Based Cloud Cover Parameterization for ICON

A promising approach to improve cloud parameterizations within climate models and thus climate projections is to use deep learning in combination with training data from storm-resolving model (SRM) simulations. The ICOsahedral Non-hydrostatic (ICON) modeling framework permits simulations ranging from numerical weather prediction to climate projections, making it an ideal target to develop neural network (NN) based parameterizations for sub-grid scale processes. Within the ICON framework, we train NN based cloud cover parameterizations with coarse-grained data based on realistic regional and global ICON SRM simulations. We set up three different types of NNs that differ in the degree of vertical locality they assume for diagnosing cloud cover from coarse-grained atmospheric state variables. The NNs accurately estimate sub-grid scale cloud cover from coarse-grained data that has similar geographical characteristics as their training data. Additionally, globally trained NNs can reproduce sub-grid scale cloud cover of the regional SRM simulation. Using the game-theory based interpretability library SHapley Additive exPlanations, we identify an overemphasis on specific humidity and cloud ice as the reason why our column-based NN cannot perfectly generalize from the global to the regional coarse-grained SRM data. The interpretability tool also helps visualize similarities and differences in feature importance between regionally and globally trained column-based NNs, and reveals a local relationship between their cloud cover predictions and the thermodynamic environment. Our results show the potential of deep learning to derive accurate yet interpretable cloud cover parameterizations from global SRMs, and suggest that neighborhood-based models may be a good compromise between accuracy and generalizability.

arXiv:2112.10889공식 코드 · Papers with Code

Surrogate Model for Shallow Water Equations Solvers with Deep Learning

Shallow water equations are the foundation of most models for flooding and river hydraulics analysis. These physics-based models are usually expensive and slow to run, thus not suitable for real-time prediction or parameter inversion. An attractive alternative is surrogate model. This work introduces an efficient, accurate, and flexible surrogate model, NN-p2p, based on deep learning and it can make point-to-point predictions on unstructured or irregular meshes. The new method was evaluated and compared against existing methods based on convolutional neural networks (CNNs), which can only make image-to-image predictions on structured or regular meshes. In NN-p2p, the input includes both spatial coordinates and boundary features that can describe the geometry of hydraulic structures, such as bridge piers. All surrogate models perform well in predicting flow around different types of piers in the training domain. However, only NN-p2p works well when spatial extrapolation is performed. The limitations of CNN-based methods are rooted in their raster-image nature which cannot capture boundary geometry and flow features exactly, which are of paramount importance to fluid dynamics. NN-p2p also has good performance in predicting flow around piers unseen by the neural network. The NN-p2p model also respects conservation laws more strictly. The application of the proposed surrogate model was demonstrated by calculating the drag coefficient $C_D$ for piers and a new linear relationship between $C_D$ and the logarithmic transformation of pier's length/width ratio was discovered.

arXiv:2112.13328공식 코드 · Papers with Code

Continuous Offline Handwriting Recognition using Deep Learning Models

Handwritten text recognition is an open problem of great interest in the area of automatic document image analysis. The transcription of handwritten content present in digitized documents is significant in analyzing historical archives or digitizing information from handwritten documents, forms, and communications. In the last years, great advances have been made in this area due to applying deep learning techniques to its resolution. This Thesis addresses the offline continuous handwritten text recognition (HTR) problem, consisting of developing algorithms and models capable of transcribing the text present in an image without the need for the text to be segmented into characters. For this purpose, we have proposed a new recognition model based on integrating two types of deep learning architectures: convolutional neural networks (CNN) and sequence-to-sequence (seq2seq) models, respectively. The convolutional component of the model is oriented to identify relevant features present in characters, and the seq2seq component builds the transcription of the text by modeling the sequential nature of the text. For the design of this new model, an extensive analysis of the capabilities of different convolutional architectures in the simplified problem of isolated character recognition has been carried out in order to identify the most suitable ones to be integrated into the continuous model. Additionally, extensive experimentation of the proposed model for the continuous problem has been carried out to determine its robustness to changes in parameterization. The generalization capacity of the model has also been validated by evaluating it on three handwritten text databases using different languages: IAM in English, RIMES in French, and Osborne in Spanish, respectively. The new proposed model provides competitive results with those obtained with other well-established methodologies.

arXiv:2201.00770공식 코드 · Papers with Code

FaceQgen: Semi-Supervised Deep Learning for Face Image Quality Assessment

In this paper we develop FaceQgen, a No-Reference Quality Assessment approach for face images based on a Generative Adversarial Network that generates a scalar quality measure related with the face recognition accuracy. FaceQgen does not require labelled quality measures for training. It is trained from scratch using the SCface database. FaceQgen applies image restoration to a face image of unknown quality, transforming it into a canonical high quality image, i.e., frontal pose, homogeneous background, etc. The quality estimation is built as the similarity between the original and the restored images, since low quality images experience bigger changes due to restoration. We compare three different numerical quality measures: a) the MSE between the original and the restored images, b) their SSIM, and c) the output score of the Discriminator of the GAN. The results demonstrate that FaceQgen's quality measures are good estimators of face recognition accuracy. Our experiments include a comparison with other quality assessment methods designed for faces and for general images, in order to position FaceQgen in the state of the art. This comparison shows that, even though FaceQgen does not surpass the best existing face quality assessment methods in terms of face recognition accuracy prediction, it achieves good enough results to demonstrate the potential of semi-supervised learning approaches for quality estimation (in particular, data-driven learning based on a single high quality image per subject), having the capacity to improve its performance in the future with adequate refinement of the model and the significant advantage over competing methods of not needing quality labels for its development. This makes FaceQgen flexible and scalable without expensive data curation.

arXiv:2112.11209공식 코드 · Papers with Code

Interpretable Knowledge Tracing: Simple and Efficient Student Modeling with Causal Relations

Intelligent Tutoring Systems have become critically important in future learning environments. Knowledge Tracing (KT) is a crucial part of that system. It is about inferring the skill mastery of students and predicting their performance to adjust the curriculum accordingly. Deep Learning-based KT models have shown significant predictive performance compared with traditional models. However, it is difficult to extract psychologically meaningful explanations from the tens of thousands of parameters in neural networks, that would relate to cognitive theory. There are several ways to achieve high accuracy in student performance prediction but diagnostic and prognostic reasoning is more critical in learning sciences. Since KT problem has few observable features (problem ID and student's correctness at each practice), we extract meaningful latent features from students' response data by using machine learning and data mining techniques. In this work, we present Interpretable Knowledge Tracing (IKT), a simple model that relies on three meaningful latent features: individual skill mastery, ability profile (learning transfer across skills), and problem difficulty. IKT's prediction of future student performance is made using a Tree-Augmented Naive Bayes Classifier (TAN), therefore its predictions are easier to explain than deep learning-based student models. IKT also shows better student performance prediction than deep learning-based student models without requiring a huge amount of parameters. We conduct ablation studies on each feature to examine their contribution to student performance prediction. Thus, IKT has great potential for providing adaptive and personalized instructions with causal reasoning in real-world educational systems.

arXiv:2201.00199공식 코드 · Papers with Code

The GatedTabTransformer. An enhanced deep learning architecture for tabular modeling

There is an increasing interest in the application of deep learning architectures to tabular data. One of the state-of-the-art solutions is TabTransformer which incorporates an attention mechanism to better track relationships between categorical features and then makes use of a standard MLP to output its final logits. In this paper we propose multiple modifications to the original TabTransformer performing better on binary classification tasks for three separate datasets with more than 1% AUROC gains. Inspired by gated MLP, linear projections are implemented in the MLP block and multiple activation functions are tested. We also evaluate the importance of specific hyper parameters during training.

arXiv:2201.00042공식 코드 · Papers with Code

Avoiding Catastrophe: Active Dendrites Enable Multi-Task Learning in Dynamic Environments

A key challenge for AI is to build embodied systems that operate in dynamically changing environments. Such systems must adapt to changing task contexts and learn continuously. Although standard deep learning systems achieve state of the art results on static benchmarks, they often struggle in dynamic scenarios. In these settings, error signals from multiple contexts can interfere with one another, ultimately leading to a phenomenon known as catastrophic forgetting. In this article we investigate biologically inspired architectures as solutions to these problems. Specifically, we show that the biophysical properties of dendrites and local inhibitory systems enable networks to dynamically restrict and route information in a context-specific manner. Our key contributions are as follows. First, we propose a novel artificial neural network architecture that incorporates active dendrites and sparse representations into the standard deep learning framework. Next, we study the performance of this architecture on two separate benchmarks requiring task-based adaptation: Meta-World, a multi-task reinforcement learning environment where a robotic agent must learn to solve a variety of manipulation tasks simultaneously; and a continual learning benchmark in which the model's prediction task changes throughout training. Analysis on both benchmarks demonstrates the emergence of overlapping but distinct and sparse subnetworks, allowing the system to fluidly learn multiple tasks with minimal forgetting. Our neural implementation marks the first time a single architecture has achieved competitive results on both multi-task and continual learning settings. Our research sheds light on how biological properties of neurons can inform deep learning systems to address dynamic scenarios that are typically impossible for traditional ANNs to solve.

arXiv:2112.11330공식 코드 · Papers with Code

PrimSeq: a deep learning-based pipeline to quantitate rehabilitation training

Stroke rehabilitation seeks to increase neuroplasticity through the repeated practice of functional motions, but may have minimal impact on recovery because of insufficient repetitions. The optimal training content and quantity are currently unknown because no practical tools exist to measure them. Here, we present PrimSeq, a pipeline to classify and count functional motions trained in stroke rehabilitation. Our approach integrates wearable sensors to capture upper-body motion, a deep learning model to predict motion sequences, and an algorithm to tally motions. The trained model accurately decomposes rehabilitation activities into component functional motions, outperforming competitive machine learning methods. PrimSeq furthermore quantifies these motions at a fraction of the time and labor costs of human experts. We demonstrate the capabilities of PrimSeq in previously unseen stroke patients with a range of upper extremity motor impairment. We expect that these advances will support the rigorous measurement required for quantitative dosing trials in stroke rehabilitation.

arXiv:2112.12409공식 코드 · Papers with Code

InstaIndoor and Multi-modal Deep Learning for Indoor Scene Recognition

Indoor scene recognition is a growing field with great potential for behaviour understanding, robot localization, and elderly monitoring, among others. In this study, we approach the task of scene recognition from a novel standpoint, using multi-modal learning and video data gathered from social media. The accessibility and variety of social media videos can provide realistic data for modern scene recognition techniques and applications. We propose a model based on fusion of transcribed speech to text and visual features, which is used for classification on a novel dataset of social media videos of indoor scenes named InstaIndoor. Our model achieves up to 70% accuracy and 0.7 F1-Score. Furthermore, we highlight the potential of our approach by benchmarking on a YouTube-8M subset of indoor scenes as well, where it achieves 74% accuracy and 0.74 F1-Score. We hope the contributions of this work pave the way to novel research in the challenging field of indoor scene recognition.

arXiv:2112.09760공식 코드 · Papers with Code

Learned Half-Quadratic Splitting Network for MR Image Reconstruction

Magnetic Resonance (MR) image reconstruction from highly undersampled $k$-space data is critical in accelerated MR imaging (MRI) techniques. In recent years, deep learning-based methods have shown great potential in this task. This paper proposes a learned half-quadratic splitting algorithm for MR image reconstruction and implements the algorithm in an unrolled deep learning network architecture. We compare the performance of our proposed method on a public cardiac MR dataset against DC-CNN and LPDNet, and our method outperforms other methods in both quantitative results and qualitative results with fewer model parameters and faster reconstruction speed. Finally, we enlarge our model to achieve superior reconstruction quality, and the improvement is $1.76$ dB and $2.74$ dB over LPDNet in peak signal-to-noise ratio on $5\times$ and $10\times$ acceleration, respectively. Code for our method is publicly available at https://github.com/hellopipu/HQS-Net.

arXiv:2112.12431공식 코드 · Papers with Code

Adaptive Modeling Against Adversarial Attacks

Adversarial training, the process of training a deep learning model with adversarial data, is one of the most successful adversarial defense methods for deep learning models. We have found that the robustness to white-box attack of an adversarially trained model can be further improved if we fine tune this model in inference stage to adapt to the adversarial input, with the extra information in it. We introduce an algorithm that "post trains" the model at inference stage between the original output class and a "neighbor" class, with existing training data. The accuracy of pre-trained Fast-FGSM CIFAR10 classifier base model against white-box projected gradient attack (PGD) can be significantly improved from 46.8% to 64.5% with our algorithm.