PyTorchKR Papers

최신 논문

총 12,307편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1906.08676공식 코드 · Papers with Code

Homogeneous Vector Capsules Enable Adaptive Gradient Descent in Convolutional Neural Networks

Capsules are the name given by Geoffrey Hinton to vector-valued neurons. Neural networks traditionally produce a scalar value for an activated neuron. Capsules, on the other hand, produce a vector of values, which Hinton argues correspond to a single, composite feature wherein the values of the components of the vectors indicate properties of the feature such as transformation or contrast. We present a new way of parameterizing and training capsules that we refer to as homogeneous vector capsules (HVCs). We demonstrate, experimentally, that altering a convolutional neural network (CNN) to use HVCs can achieve superior classification accuracy without increasing the number of parameters or operations in its architecture as compared to a CNN using a single final fully connected layer. Additionally, the introduction of HVCs enables the use of adaptive gradient descent, reducing the dependence a model's achievable accuracy has on the finely tuned hyperparameters of a non-adaptive optimizer. We demonstrate our method and results using two neural network architectures. First, a very simple monolithic CNN that when using HVCs achieved a 63% improvement in top-1 classification accuracy and a 35% improvement in top-5 classification accuracy over the baseline architecture. Second, with the CNN architecture referred to as Inception v3 that achieved similar accuracies both with and without HVCs. Additionally, the simple monolithic CNN when using HVCs showed no overfitting after more than 300 epochs whereas the baseline showed overfitting after 30 epochs. We use the ImageNet ILSVRC 2012 classification challenge dataset with both networks.

arXiv:1906.08615

Zero-shot Learning and Knowledge Transfer in Music Classification and Tagging

Music classification and tagging is conducted through categorical supervised learning with a fixed set of labels. In principle, this cannot make predictions on unseen labels. Zero-shot learning is an approach to solve the problem by using side information about the semantic labels. We recently investigated this concept of zero-shot learning in music classification and tagging task by projecting both audio and label space on a single semantic space. In this work, we extend the work to verify the generalization ability of zero-shot learning model by conducting knowledge transfer to different music corpora.

arXiv:1906.08541공식 코드 · Papers with Code

Regional based query in graph active learning

Graph convolution networks (GCN) have emerged as the leading method to classify node classes in networks, and have reached the highest accuracy in multiple node classification tasks. In the absence of available tagged samples, active learning methods have been developed to obtain the highest accuracy using the minimal number of queries to an oracle. The current best active learning methods use the sample class uncertainty as selection criteria. However, in graph based classification, the class of each node is often related to the class of its neighbors. As such, the uncertainty in the class of a node's neighbor may be a more appropriate selection criterion. We here propose two such criteria, one extending the classical uncertainty measure, and the other extending the page-rank algorithm. We show that the latter is optimal when the fraction of tagged nodes is low, and when this fraction grows to one over the average degree, the regional uncertainty performs better than all existing methods. While we have tested this methods on graphs, such methods can be extended to any classification problem, where a distance metrics can be defined between the input samples. All the code used can be accessed at : https://github.com/louzounlab/graph-al All the datasets used can be accessed at : https://github.com/louzounlab/DataSets

arXiv:1906.08244

Predicting Patent Citations to measure Economic Impact of Scholarly Research

A crucial goal of funding research and development has always been to advance economic development. On this basis, a consider-able body of research undertaken with the purpose of determining what exactly constitutes economic impact and how to accurately measure that impact has been published. Numerous indicators have been used to measure economic impact, although no single indicator has been widely adapted. Based on patent data collected from Altmetric we predict patent citations through various social media features using several classification models. Patents citing a research paper implies the potential it has for direct application inits field. These predictions can be utilized by researchers in deter-mining the practical applications for their work when applying for patents.

arXiv:1906.08045

Speech Recognition With No Speech Or With Noisy Speech Beyond English

In this paper we demonstrate continuous noisy speech recognition using connectionist temporal classification (CTC) model on limited Chinese vocabulary using electroencephalography (EEG) features with no speech signal as input and we further demonstrate single CTC model based continuous noisy speech recognition on limited joint English and Chinese vocabulary using EEG features with no speech signal as input. We demonstrate our results using various EEG feature sets recently introduced in [1] as well as we propose a new deep learning architecture in this paper which can perform continuous speech recognition using raw EEG signals on limited joint English and Chinese vocabulary.

arXiv:1906.07775

Quantifying and Leveraging Classification Uncertainty for Chest Radiograph Assessment

The interpretation of chest radiographs is an essential task for the detection of thoracic diseases and abnormalities. However, it is a challenging problem with high inter-rater variability and inherent ambiguity due to inconclusive evidence in the data, limited data quality or subjective definitions of disease appearance. Current deep learning solutions for chest radiograph abnormality classification are typically limited to providing probabilistic predictions, relying on the capacity of learning models to adapt to the high degree of label noise and become robust to the enumerated causal factors. In practice, however, this leads to overconfident systems with poor generalization on unseen data. To account for this, we propose an automatic system that learns not only the probabilistic estimate on the presence of an abnormality, but also an explicit uncertainty measure which captures the confidence of the system in the predicted output. We argue that explicitly learning the classification uncertainty as an orthogonal measure to the predicted output, is essential to account for the inherent variability characteristic of this data. Experiments were conducted on two datasets of chest radiographs of over 85,000 patients. Sample rejection based on the predicted uncertainty can significantly improve the ROC-AUC, e.g., by 8% to 0.91 with an expected rejection rate of under 25%. Eliminating training samples using uncertainty-driven bootstrapping, enables a significant increase in robustness and accuracy. In addition, we present a multi-reader study showing that the predictive uncertainty is indicative of reader errors.

arXiv:1906.07771공식 코드 · Papers with Code

Crop Lodging Prediction from UAV-Acquired Images of Wheat and Canola using a DCNN Augmented with Handcrafted Texture Features

Lodging, the permanent bending over of food crops, leads to poor plant growth and development. Consequently, lodging results in reduced crop quality, lowers crop yield, and makes harvesting difficult. Plant breeders routinely evaluate several thousand breeding lines, and therefore, automatic lodging detection and prediction is of great value aid in selection. In this paper, we propose a deep convolutional neural network (DCNN) architecture for lodging classification using five spectral channel orthomosaic images from canola and wheat breeding trials. Also, using transfer learning, we trained 10 lodging detection models using well-established deep convolutional neural network architectures. Our proposed model outperforms the state-of-the-art lodging detection methods in the literature that use only handcrafted features. In comparison to 10 DCNN lodging detection models, our proposed model achieves comparable results while having a substantially lower number of parameters. This makes the proposed model suitable for applications such as real-time classification using inexpensive hardware for high-throughput phenotyping pipelines. The GitHub repository at https://github.com/FarhadMaleki/LodgedNet contains code and models.

arXiv:1906.07749

A Framework for Parallelizing OWL Classification in Description Logic Reasoners

In this paper we report on a black-box approach to parallelize existing description logic (DL) reasoners for the Web Ontology Language (OWL). We focus on OWL ontology classification, which is an important inference service and supported by every major OWL/DL reasoner. We propose a flexible parallel framework which can be applied to existing OWL reasoners in order to speed up their classification process. In order to test its performance, we evaluated our framework by parallelizing major OWL reasoners for concept classification. In comparison to the selected black-box reasoner our results demonstrate that the wall clock time of ontology classification can be improved by one order of magnitude for most real-world ontologies.

arXiv:1906.08215공식 코드 · Papers with Code

Bayesian Learning from Sequential Data using Gaussian Processes with Signature Covariances

We develop a Bayesian approach to learning from sequential data by using Gaussian processes (GPs) with so-called signature kernels as covariance functions. This allows to make sequences of different length comparable and to rely on strong theoretical results from stochastic analysis. Signatures capture sequential structure with tensors that can scale unfavourably in sequence length and state space dimension. To deal with this, we introduce a sparse variational approach with inducing tensors. We then combine the resulting GP with LSTMs and GRUs to build larger models that leverage the strengths of each of these approaches and benchmark the resulting GPs on multivariate time series (TS) classification datasets. Code available at https://github.com/tgcsaba/GPSig.

arXiv:1906.08129

Efficient Set-Valued Prediction in Multi-Class Classification

In cases of uncertainty, a multi-class classifier preferably returns a set of candidate classes instead of predicting a single class label with little guarantee. More precisely, the classifier should strive for an optimal balance between the correctness (the true class is among the candidates) and the precision (the candidates are not too many) of its prediction. We formalize this problem within a general decision-theoretic framework that unifies most of the existing work in this area. In this framework, uncertainty is quantified in terms of conditional class probabilities, and the quality of a predicted set is measured in terms of a utility function. We then address the problem of finding the Bayes-optimal prediction, i.e., the subset of class labels with highest expected utility. For this problem, which is computationally challenging as there are exponentially (in the number of classes) many predictions to choose from, we propose efficient algorithms that can be applied to a broad family of utility functions. Our theoretical results are complemented by experimental studies, in which we analyze the proposed algorithms in terms of predictive accuracy and runtime efficiency.

arXiv:1906.08101공식 코드 · Papers with Code

Pre-Training with Whole Word Masking for Chinese BERT

Bidirectional Encoder Representations from Transformers (BERT) has shown marvelous improvements across various NLP tasks, and its consecutive variants have been proposed to further improve the performance of the pre-trained language models. In this paper, we aim to first introduce the whole word masking (wwm) strategy for Chinese BERT, along with a series of Chinese pre-trained language models. Then we also propose a simple but effective model called MacBERT, which improves upon RoBERTa in several ways. Especially, we propose a new masking strategy called MLM as correction (Mac). To demonstrate the effectiveness of these models, we create a series of Chinese pre-trained language models as our baselines, including BERT, RoBERTa, ELECTRA, RBT, etc. We carried out extensive experiments on ten Chinese NLP tasks to evaluate the created Chinese pre-trained language models as well as the proposed MacBERT. Experimental results show that MacBERT could achieve state-of-the-art performances on many NLP tasks, and we also ablate details with several findings that may help future research. We open-source our pre-trained language models for further facilitating our research community. Resources are available: https://github.com/ymcui/Chinese-BERT-wwm

arXiv:1906.07997

Cloud-based Image Classification Service Is Not Robust To Simple Transformations: A Forgotten Battlefield

Many recent works demonstrated that Deep Learning models are vulnerable to adversarial examples.Fortunately, generating adversarial examples usually requires white-box access to the victim model, and the attacker can only access the APIs opened by cloud platforms. Thus, keeping models in the cloud can usually give a (false) sense of security.Unfortunately, cloud-based image classification service is not robust to simple transformations such as Gaussian Noise, Salt-and-Pepper Noise, Rotation and Monochromatization. In this paper,(1) we propose one novel attack method called Image Fusion(IF) attack, which achieve a high bypass rate,can be implemented only with OpenCV and is difficult to defend; and (2) we make the first attempt to conduct an extensive empirical study of Simple Transformation (ST) attacks against real-world cloud-based classification services. Through evaluations on four popular cloud platforms including Amazon, Google, Microsoft, Clarifai, we demonstrate that ST attack has a success rate of approximately 100% except Amazon approximately 50%, IF attack have a success rate over 98% among different classification services. (3) We discuss the possible defenses to address these security challenges.Experiments show that our defense technology can effectively defend known ST attacks.

arXiv:1906.07934

A simple and effective postprocessing method for image classification

Whether it is computer vision, natural language processing or speech recognition, the essence of these applications is to obtain powerful feature representations that make downstream applications completion more efficient. Taking image recognition as an example, whether it is hand-crafted low-level feature representation or feature representation extracted by a convolutional neural networks(CNNs), the goal is to extract features that better represent image features, thereby improving classification accuracy. However, we observed that image feature representations share a large common vector and a few top dominating directions. To address this problems, we propose a simple but effective postprocessing method to render off-the-shelf feature representations even stronger by eliminating the common mean vector from off-the-shelf feature representations. The postprocessing is empirically validated on a variety of datasets and feature extraction methods.such as VGG, LBP, and HOG. Some experiments show that the features that have been post-processed by postprocessing algorithm can get better results than original ones.

arXiv:1906.07912공식 코드 · Papers with Code

ViP: Virtual Pooling for Accelerating CNN-based Image Classification and Object Detection

In recent years, Convolutional Neural Networks (CNNs) have shown superior capability in visual learning tasks. While accuracy-wise CNNs provide unprecedented performance, they are also known to be computationally intensive and energy demanding for modern computer systems. In this paper, we propose Virtual Pooling (ViP), a model-level approach to improve speed and energy consumption of CNN-based image classification and object detection tasks, with a provable error bound. We show the efficacy of ViP through experiments on four CNN models, three representative datasets, both desktop and mobile platforms, and two visual learning tasks, i.e., image classification and object detection. For example, ViP delivers 2.1x speedup with less than 1.5% accuracy degradation in ImageNet classification on VGG-16, and 1.8x speedup with 0.025 mAP degradation in PASCAL VOC object detection with Faster-RCNN. ViP also reduces mobile GPU and CPU energy consumption by up to 55% and 70%, respectively. As a complementary method to existing acceleration approaches, ViP achieves 1.9x speedup on ThiNet leading to a combined speedup of 5.23x on VGG-16. Furthermore, ViP provides a knob for machine learning practitioners to generate a set of CNN models with varying trade-offs between system speed/energy consumption and accuracy to better accommodate the requirements of their tasks. Code is available at https://github.com/cmu-enyac/VirtualPooling.

arXiv:1906.07899

Brain correlates of task-load and dementia elucidation with tensor machine learning using oddball BCI paradigm

Dementia in the elderly has recently become the most usual cause of cognitive decline. The proliferation of dementia cases in aging societies creates a remarkable economic as well as medical problems in many communities worldwide. A recently published report by The World Health Organization (WHO) estimates that about 47 million people are suffering from dementia-related neurocognitive declines worldwide. The number of dementia cases is predicted by 2050 to triple, which requires the creation of an AI-based technology application to support interventions with early screening for subsequent mental wellbeing checking as well as preservation with digital-pharma (the so-called beyond a pill) therapeutical approaches. We present an attempt and exploratory results of brain signal (EEG) classification to establish digital biomarkers for dementia stage elucidation. We discuss a comparison of various machine learning approaches for automatic event-related potentials (ERPs) classification of a high and low task-load sound stimulus recognition. These ERPs are similar to those in dementia. The proposed winning method using tensor-based machine learning in a deep fully connected neural network setting is a step forward to develop AI-based approaches for a subsequent application for subjective- and mild-cognitive impairment (SCI and MCI) diagnostics.

arXiv:1906.07882

Semi-supervised Logistic Learning Based on Exponential Tilt Mixture Models

Consider semi-supervised learning for classification, where both labeled and unlabeled data are available for training. The goal is to exploit both datasets to achieve higher prediction accuracy than just using labeled data alone. We develop a semi-supervised logistic learning method based on exponential tilt mixture models, by extending a statistical equivalence between logistic regression and exponential tilt modeling. We study maximum nonparametric likelihood estimation and derive novel objective functions which are shown to be Fisher consistent. We also propose regularized estimation and construct simple and highly interpretable EM algorithms. Finally, we present numerical results which demonstrate the advantage of the proposed methods compared with existing methods.

arXiv:1906.07599

LTG-Oslo Hierarchical Multi-task Network: The importance of negation for document-level sentiment in Spanish

This paper details LTG-Oslo team's participation in the sentiment track of the NEGES 2019 evaluation campaign. We participated in the task with a hierarchical multi-task network, which used shared lower-layers in a deep BiLSTM to predict negation, while the higher layers were dedicated to predicting document-level sentiment. The multi-task component shows promise as a way to incorporate information on negation into deep neural sentiment classifiers, despite the fact that the absolute results on the test set were relatively low for a binary classification task.

arXiv:1906.07525

Mimicking Human Process: Text Representation via Latent Semantic Clustering for Classification

Considering that words with different characteristic in the text have different importance for classification, grouping them together separately can strengthen the semantic expression of each part. Thus we propose a new text representation scheme by clustering words according to their latent semantics and composing them together to get a set of cluster vectors, which are then concatenated as the final text representation. Evaluation on five classification benchmarks proves the effectiveness of our method. We further conduct visualization analysis showing statistical clustering results and verifying the validity of our motivation.

arXiv:1906.07282공식 코드 · Papers with Code

The Cells Out of Sample (COOS) dataset and benchmarks for measuring out-of-sample generalization of image classifiers

Understanding if classifiers generalize to out-of-sample datasets is a central problem in machine learning. Microscopy images provide a standardized way to measure the generalization capacity of image classifiers, as we can image the same classes of objects under increasingly divergent, but controlled factors of variation. We created a public dataset of 132,209 images of mouse cells, COOS-7 (Cells Out Of Sample 7-Class). COOS-7 provides a classification setting where four test datasets have increasing degrees of covariate shift: some images are random subsets of the training data, while others are from experiments reproduced months later and imaged by different instruments. We benchmarked a range of classification models using different representations, including transferred neural network features, end-to-end classification with a supervised deep CNN, and features from a self-supervised CNN. While most classifiers perform well on test datasets similar to the training dataset, all classifiers failed to generalize their performance to datasets with greater covariate shifts. These baselines highlight the challenges of covariate shifts in image data, and establish metrics for improving the generalization capacity of image classifiers.

arXiv:1906.07273

Coherent and Controllable Outfit Generation

When thinking about dressing oneself, people often have a theme in mind whether they're going to a tropical getaway or wish to appear attractive at a cocktail party. A useful outfit generation system should come up with clothing items that are compatible while matching a theme specified by the user. Existing methods use item-wise compatibility between products but lack an effective way to enforce a global constraint (e.g., style, occasion). We introduce a method that generates outfits whose items match a theme described by a text query. Our method uses text and image embeddings to represent fashion items. We learn a multimodal embedding where the image representation for an item is close to its text representation, and use this embedding to measure item-query coherence. We then use a discriminator to compute compatibility between fashion items. This strategy yields a compatibility prediction method that meets or exceeds the state of the art. Our method combines item-item compatibility and item-query coherence to construct an outfit whose items are (a) close to the query and (b) compatible with one another. Quantitative evaluation shows that the items in our outfits are tightly clustered compared to standard outfits. Furthermore, outfits produced by similar queries are close to one another, and outfits produced by very different queries are far apart. Qualitative evaluation shows that our method responds well to queries. A user study suggests that people understand the match between the queries and the outfits produced by our method.