PyTorchKR Papers

최신 논문

총 12,350편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1807.01035

Deep Neural Object Analysis by Interactive Auditory Exploration with a Humanoid Robot

We present a novel approach for interactive auditory object analysis with a humanoid robot. The robot elicits sensory information by physically shaking visually indistinguishable plastic capsules. It gathers the resulting audio signals from microphones that are embedded into the robotic ears. A neural network architecture learns from these signals to analyze properties of the contents of the containers. Specifically, we evaluate the material classification and weight prediction accuracy and demonstrate that the framework is fairly robust to acoustic real-world noise.

arXiv:1807.01026

Deep Architectures and Ensembles for Semantic Video Classification

This work addresses the problem of accurate semantic labelling of short videos. To this end, a multitude of different deep nets, ranging from traditional recurrent neural networks (LSTM, GRU), temporal agnostic networks (FV,VLAD,BoW), fully connected neural networks mid-stage AV fusion and others. Additionally, we also propose a residual architecture-based DNN for video classification, with state-of-the art classification performance at significantly reduced complexity. Furthermore, we propose four new approaches to diversity-driven multi-net ensembling, one based on fast correlation measure and three incorporating a DNN-based combiner. We show that significant performance gains can be achieved by ensembling diverse nets and we investigate factors contributing to high diversity. Based on the extensive YouTube8M dataset, we provide an in-depth evaluation and analysis of their behaviour. We show that the performance of the ensemble is state-of-the-art achieving the highest accuracy on the YouTube-8M Kaggle test data. The performance of the ensemble of classifiers was also evaluated on the HMDB51 and UCF101 datasets, and show that the resulting method achieves comparable accuracy with state-of-the-art methods using similar input features.

arXiv:1807.00958

Iterative Attention Mining for Weakly Supervised Thoracic Disease Pattern Localization in Chest X-Rays

Given image labels as the only supervisory signal, we focus on harvesting, or mining, thoracic disease localizations from chest X-ray images. Harvesting such localizations from existing datasets allows for the creation of improved data sources for computer-aided diagnosis and retrospective analyses. We train a convolutional neural network (CNN) for image classification and propose an attention mining (AM) strategy to improve the model's sensitivity or saliency to disease patterns. The intuition of AM is that once the most salient disease area is blocked or hidden from the CNN model, it will pay attention to alternative image regions, while still attempting to make correct predictions. However, the model requires to be properly constrained during AM, otherwise, it may overfit to uncorrelated image parts and forget the valuable knowledge that it has learned from the original image classification task. To alleviate such side effects, we then design a knowledge preservation (KP) loss, which minimizes the discrepancy between responses for X-ray images from the original and the updated networks. Furthermore, we modify the CNN model to include multi-scale aggregation (MSA), improving its localization ability on small-scale disease findings, e.g., lung nodules. We experimentally validate our method on the publicly-available ChestX-ray14 dataset, outperforming a class activation map (CAM)-based approach, and demonstrating the value of our novel framework for mining disease locations.

arXiv:1807.00211

Advanced Methods for the Optical Quality Assurance of Silicon Sensors

We describe a setup for optical quality assurance of silicon microstrip sensors. Pattern recognition algorithms were developed to analyze microscopic scans of the sensors for defects. It is shown that the software has a recognition and classification rate of $>$~90\% for defects like scratches, shorts, broken metal lines etc. We have demonstrated that advanced image processing based on neural network techniques is able to further improve the recognition and defect classification rate.

arXiv:1807.00141

Fractional Wavelet Scattering Network and Applications

Objective: The present study introduces a fractional wavelet scattering network (FrScatNet), which is a generalized translation invariant version of the classical wavelet scattering network (ScatNet). Methods: In our approach, the FrScatNet is constructed based on the fractional wavelet transform (FRWT). The fractional scattering coefficients are iteratively computed using FRWTs and modulus operators. The feature vectors constructed by fractional scattering coefficients are usually used for signal classification. In this work, an application example of FrScatNet is provided in order to assess its performance on pathological images. Firstly, the FrScatNet extracts feature vectors from patches of the original histological images under different orders. Then we classify those patches into target (benign or malignant) and background groups. And the FrScatNet property is analyzed by comparing error rates computed from different fractional orders respectively. Based on the above pathological image classification, a gland segmentation algorithm is proposed by combining the boundary information and the gland location. Results: The error rates for different fractional orders of FrScatNet are examined and show that the classification accuracy is significantly improved in fractional scattering domain. We also compare the FrScatNet based gland segmentation method with those proposed in the 2015 MICCAI Gland Segmentation Challenge and our method achieves comparable results. Conclusion: The FrScatNet is shown to achieve accurate and robust results. More stable and discriminative fractional scattering coefficients are obtained by the FrScatNet in this work. Significance: The added fractional order parameter is able to analyze the image in the fractional scattering domain.

arXiv:1807.00094

Classification of lung nodules in CT images based on Wasserstein distance in differential geometry

Lung nodules are commonly detected in screening for patients with a risk for lung cancer. Though the status of large nodules can be easily diagnosed by fine needle biopsy or bronchoscopy, small nodules are often difficult to classify on computed tomography (CT). Recent works have shown that shape analysis of lung nodules can be used to differentiate benign lesions from malignant ones, though existing methods are limited in their sensitivity and specificity. In this work we introduced a new 3D shape analysis within the framework of differential geometry to calculate the Wasserstein distance between benign and malignant lung nodules to derive an accurate classification scheme. The Wasserstein distance between the nodules is calculated based on our new spherical optimal mass transport, this new algorithm works directly on sphere by using spherical metric, which is much more accurate and efficient than previous methods. In the process of deformation, the area-distortion factor gives a probability measure on the unit sphere, which forms the Wasserstein space. From known cases of benign and malignant lung nodules, we can calculate a unique optimal mass transport map between their correspondingly deformed Wasserstein spaces. This transportation cost defines the Wasserstein distance between them and can be used to classify new lung nodules into either the benign or malignant class. To the best of our knowledge, this is the first work that utilizes Wasserstein distance for lung nodule classification. The advantages of Wasserstein distance are it is invariant under rigid motions and scalings, thus it intrinsically measures shape distance even when the underlying shapes are of high complexity, making it well suited to classify lung nodules as they have different sizes, orientations, and appearances.

arXiv:1807.00083

Topology classification with deep learning to improve real-time event selection at the LHC

We show how event topology classification based on deep learning could be used to improve the purity of data samples selected in real time at at the Large Hadron Collider. We consider different data representations, on which different kinds of multi-class classifiers are trained. Both raw data and high-level features are utilized. In the considered examples, a filter based on the classifier's score can be trained to retain ~99% of the interesting events and reduce the false-positive rate by as much as one order of magnitude for certain background processes. By operating such a filter as part of the online event selection infrastructure of the LHC experiments, one could benefit from a more flexible and inclusive selection strategy while reducing the amount of downstream resources wasted in processing false positives. The saved resources could be translated into a reduction of the detector operation cost or into an effective increase of storage and processing capabilities, which could be reinvested to extend the physics reach of the LHC experiments.

arXiv:1807.00072

Joint Learning of Domain Classification and Out-of-Domain Detection with Dynamic Class Weighting for Satisficing False Acceptance Rates

In domain classification for spoken dialog systems, correct detection of out-of-domain (OOD) utterances is crucial because it reduces confusion and unnecessary interaction costs between users and the systems. Previous work usually utilizes OOD detectors that are trained separately from in-domain (IND) classifiers, and confidence thresholding for OOD detection given target evaluation scores. In this paper, we introduce a neural joint learning model for domain classification and OOD detection, where dynamic class weighting is used during the model training to satisfice a given OOD false acceptance rate (FAR) while maximizing the domain classification accuracy. Evaluating on two domain classification tasks for the utterances from a large spoken dialogue system, we show that our approach significantly improves the domain classification performance with satisficing given target FARs.

arXiv:1807.00053공식 코드 · Papers with Code

Task-Driven Convolutional Recurrent Models of the Visual System

Feed-forward convolutional neural networks (CNNs) are currently state-of-the-art for object classification tasks such as ImageNet. Further, they are quantitatively accurate models of temporally-averaged responses of neurons in the primate brain's visual system. However, biological visual systems have two ubiquitous architectural features not shared with typical CNNs: local recurrence within cortical areas, and long-range feedback from downstream areas to upstream areas. Here we explored the role of recurrence in improving classification performance. We found that standard forms of recurrence (vanilla RNNs and LSTMs) do not perform well within deep CNNs on the ImageNet task. In contrast, novel cells that incorporated two structural features, bypassing and gating, were able to boost task accuracy substantially. We extended these design principles in an automated search over thousands of model architectures, which identified novel local recurrent cells and long-range feedback connections useful for object recognition. Moreover, these task-optimized ConvRNNs matched the dynamics of neural activity in the primate visual system better than feedforward networks, suggesting a role for the brain's recurrent connections in performing difficult visual behaviors.

arXiv:1807.00804

Classifying Data with Local Hamiltonians

The goal of this work is to define a notion of a quantum neural network to classify data, which exploits the low energy spectrum of a local Hamiltonian. As a concrete application, we build a binary classifier, train it on some actual data and then test its performance on a simple classification task. More specifically, we use Microsoft's quantum simulator, Liquid, to construct local Hamiltonians that can encode trained classifier functions in their ground space, and which can be probed by measuring the overlap with test states corresponding to the data to be classified. To obtain such a classifier Hamiltonian, we further propose a training scheme based on quantum annealing which is completely closed-off to the environment and which does not depend on external measurements until the very end, avoiding unnecessary decoherence during the annealing procedure. For a network of size n, the trained network can be stored as a list of O(n) coupling strengths. We address the question of which interactions are most suitable for a given classification task, and develop a qubit-saving optimization for the training procedure on a simulated annealing device. Furthermore, a small neural network to classify colors into red vs. blue is trained and tested, and benchmarked against the annealing parameters.

arXiv:1807.00578

Classifying neuromorphic data using a deep learning framework for image classification

In the field of artificial intelligence, neuromorphic computing has been around for several decades. Deep learning has however made much recent progress such that it consistently outperforms neuromorphic learning algorithms in classification tasks in terms of accuracy. Specifically in the field of image classification, neuromorphic computing has been traditionally using either the temporal or rate code for encoding static images in datasets into spike trains. It is only till recently, that neuromorphic vision sensors are widely used by the neuromorphic research community, and provides an alternative to such encoding methods. Since then, several neuromorphic datasets as obtained by applying such sensors on image datasets (e.g. the neuromorphic CALTECH 101) have been introduced. These data are encoded in spike trains and hence seem ideal for benchmarking of neuromorphic learning algorithms. Specifically, we train a deep learning framework used for image classification on the CALTECH 101 and a collapsed version of the neuromorphic CALTECH 101 datasets. We obtained an accuracy of 91.66% and 78.01% for the CALTECH 101 and neuromorphic CALTECH 101 datasets respectively. For CALTECH 101, our accuracy is close to the best reported accuracy, while for neuromorphic CALTECH 101, it outperforms the last best reported accuracy by over 10%. This raises the question of the suitability of such datasets as benchmarks for neuromorphic learning algorithms.

arXiv:1807.00488

A Simple but Effective Classification Model for Grammatical Error Correction

We treat grammatical error correction (GEC) as a classification problem in this study, where for different types of errors, a target word is identified, and the classifier predicts the correct word form from a set of possible choices. We propose a novel neural network based feature representation and classification model, trained using large text corpora without human annotations. Specifically we use RNNs with attention to represent both the left and right context of a target word. All feature embeddings are learned jointly in an end-to-end fashion. Experimental results show that our novel approach outperforms other classifier methods on the CoNLL-2014 test set (F0.5 45.05%). Our model is simple but effective, and is suitable for industrial production.

arXiv:1807.00480

FATE: Fast and Accurate Timing Error Prediction Framework for Low Power DNN Accelerator Design

Deep neural networks (DNN) are increasingly being accelerated on application-specific hardware such as the Google TPU designed especially for deep learning. Timing speculation is a promising approach to further increase the energy efficiency of DNN accelerators. Architectural exploration for timing speculation requires detailed gate-level timing simulations that can be time-consuming for large DNNs that execute millions of multiply-and-accumulate (MAC) operations. In this paper we propose FATE, a new methodology for fast and accurate timing simulations of DNN accelerators like the Google TPU. FATE proposes two novel ideas: (i) DelayNet, a DNN based timing model for MAC units; and (ii) a statistical sampling methodology that reduces the number of MAC operations for which timing simulations are performed. We show that FATE results in between 8 times-58 times speed-up in timing simulations, while introducing less than 2% error in classification accuracy estimates. We demonstrate the use of FATE by comparing to conventional DNN accelerator that uses 2's complement (2C) arithmetic with an alternative implementation that uses signed magnitude representations (SMR). We show that that the SMR implementation provides 18% more energy savings for the same classification accuracy than 2C, a result that might be of independent interest.

arXiv:1807.00458

Adversarial Perturbations Against Real-Time Video Classification Systems

Recent research has demonstrated the brittleness of machine learning systems to adversarial perturbations. However, the studies have been mostly limited to perturbations on images and more generally, classification that does not deal with temporally varying inputs. In this paper we ask "Are adversarial perturbations possible in real-time video classification systems and if so, what properties must they satisfy?" Such systems find application in surveillance applications, smart vehicles, and smart elderly care and thus, misclassification could be particularly harmful (e.g., a mishap at an elderly care facility may be missed). We show that accounting for temporal structure is key to generating adversarial examples in such systems. We exploit recent advances in generative adversarial network (GAN) architectures to account for temporal correlations and generate adversarial samples that can cause misclassification rates of over 80% for targeted activities. More importantly, the samples also leave other activities largely unaffected making them extremely stealthy. Finally, we also surprisingly find that in many scenarios, the same perturbation can be applied to every frame in a video clip that makes the adversary's ability to achieve misclassification relatively easy.

arXiv:1807.00453

Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision

We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.

arXiv:1807.00340

Towards Adversarial Training with Moderate Performance Improvement for Neural Network Classification

It has been demonstrated that deep neural networks are prone to noisy examples particular adversarial samples during inference process. The gap between robust deep learning systems in real world applications and vulnerable neural networks is still large. Current adversarial training strategies improve the robustness against adversarial samples. However, these methods lead to accuracy reduction when the input examples are clean thus hinders the practicability. In this paper, we investigate an approach that protects the neural network classification from the adversarial samples and improves its accuracy when the input examples are clean. We demonstrate the versatility and effectiveness of our proposed approach on a variety of different networks and datasets.

arXiv:1807.00284

Autonomous Deep Learning: A Genetic DCNN Designer for Image Classification

Recent years have witnessed the breakthrough success of deep convolutional neural networks (DCNNs) in image classification and other vision applications. Although freeing users from the troublesome handcrafted feature extraction by providing a uniform feature extraction-classification framework, DCNNs still require a handcrafted design of their architectures. In this paper, we propose the genetic DCNN designer, an autonomous learning algorithm can generate a DCNN architecture automatically based on the data available for a specific image classification problem. We first partition a DCNN into multiple stacked meta convolutional blocks and fully connected blocks, each containing the operations of convolution, pooling, fully connection, batch normalization, activation and drop out, and thus convert the architecture into an integer vector. Then, we use refined evolutionary operations, including selection, mutation and crossover to evolve a population of DCNN architectures. Our results on the MNIST, Fashion-MNIST, EMNISTDigit, EMNIST-Letter, CIFAR10 and CIFAR100 datasets suggest that the proposed genetic DCNN designer is able to produce automatically DCNN architectures, whose performance is comparable to, if not better than, that of stateof- the-art DCNN models

arXiv:1806.11146

Adversarial Reprogramming of Neural Networks

Deep neural networks are susceptible to \emph{adversarial} attacks. In computer vision, well-crafted perturbations to images can cause neural networks to make mistakes such as confusing a cat with a computer. Previous adversarial attacks have been designed to degrade performance of models or cause machine learning models to produce specific outputs chosen ahead of time by the attacker. We introduce attacks that instead {\em reprogram} the target model to perform a task chosen by the attacker---without the attacker needing to specify or compute the desired output for each test-time input. This attack finds a single adversarial perturbation, that can be added to all test-time inputs to a machine learning model in order to cause the model to perform a task chosen by the adversary---even if the model was not trained to do this task. These perturbations can thus be considered a program for the new task. We demonstrate adversarial reprogramming on six ImageNet classification models, repurposing these models to perform a counting task, as well as classification tasks: classification of MNIST and CIFAR-10 examples presented as inputs to the ImageNet model.

arXiv:1806.11391공식 코드 · Papers with Code

A Comparative Study of Distributional and Symbolic Paradigms for Relational Learning

Many real-world domains can be expressed as graphs and, more generally, as multi-relational knowledge graphs. Though reasoning and learning with knowledge graphs has traditionally been addressed by symbolic approaches, recent methods in (deep) representation learning has shown promising results for specialized tasks such as knowledge base completion. These approaches abandon the traditional symbolic paradigm by replacing symbols with vectors in Euclidean space. With few exceptions, symbolic and distributional approaches are explored in different communities and little is known about their respective strengths and weaknesses. In this work, we compare representation learning and relational learning on various relational classification and clustering tasks and analyse the complexity of the rules used implicitly by these approaches. Preliminary results reveal possible indicators that could help in choosing one approach over the other for particular knowledge graphs.

arXiv:1806.11186

A New Angle on L2 Regularization

Imagine two high-dimensional clusters and a hyperplane separating them. Consider in particular the angle between: the direction joining the two clusters' centroids and the normal to the hyperplane. In linear classification, this angle depends on the level of L2 regularization used. Can you explain why?