PyTorchKR Papers

최신 논문

총 4,526편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1612.06139

Neural Machine Translation from Simplified Translations

Text simplification aims at reducing the lexical, grammatical and structural complexity of a text while keeping the same meaning. In the context of machine translation, we introduce the idea of simplified translations in order to boost the learning ability of deep neural translation models. We conduct preliminary experiments showing that translation complexity is actually reduced in a translation of a source bi-text compared to the target reference of the bi-text while using a neural machine translation (NMT) system learned on the exact same bi-text. Based on knowledge distillation idea, we then train an NMT system using the simplified bi-text, and show that it outperforms the initial system that was built over the reference data set. Performance is further boosted when both reference and automatic translations are used to learn the network. We perform an elementary analysis of the translated corpus and report accuracy results of the proposed approach on English-to-French and English-to-German translation tasks.

arXiv:1608.00892

Knowledge Distillation for Small-footprint Highway Networks

Deep learning has significantly advanced state-of-the-art of speech recognition in the past few years. However, compared to conventional Gaussian mixture acoustic models, neural network models are usually much larger, and are therefore not very deployable in embedded devices. Previously, we investigated a compact highway deep neural network (HDNN) for acoustic modelling, which is a type of depth-gated feedforward neural network. We have shown that HDNN-based acoustic models can achieve comparable recognition accuracy with much smaller number of model parameters compared to plain deep neural network (DNN) acoustic models. In this paper, we push the boundary further by leveraging on the knowledge distillation technique that is also known as {\it teacher-student} training, i.e., we train the compact HDNN model with the supervision of a high accuracy cumbersome model. Furthermore, we also investigate sequence training and adaptation in the context of teacher-student training. Our experiments were performed on the AMI meeting speech recognition corpus. With this technique, we significantly improved the recognition accuracy of the HDNN acoustic model with less than 0.8 million parameters, and narrowed the gap between this model and the plain DNN with 30 million parameters.

arXiv:1609.05695

A scalable convolutional neural network for task-specified scenarios via knowledge distillation

In this paper, we explore the redundancy in convolutional neural network, which scales with the complexity of vision tasks. Considering that many front-end visual systems are interested in only a limited range of visual targets, the removing of task-specified network redundancy can promote a wide range of potential applications. We propose a task-specified knowledge distillation algorithm to derive a simplified model with pre-set computation cost and minimized accuracy loss, which suits the resource constraint front-end systems well. Experiments on the MNIST and CIFAR10 datasets demonstrate the feasibility of the proposed approach as well as the existence of task-specified redundancy.

arXiv:1702.02052

Knowledge Adaptation: Teaching to Adapt

Domain adaptation is crucial in many real-world applications where the distribution of the training data differs from the distribution of the test data. Previous Deep Learning-based approaches to domain adaptation need to be trained jointly on source and target domain data and are therefore unappealing in scenarios where models need to be adapted to a large number of domains or where a domain is evolving, e.g. spam detection where attackers continuously change their tactics. To fill this gap, we propose Knowledge Adaptation, an extension of Knowledge Distillation (Bucilua et al., 2006; Hinton et al., 2015) to the domain adaptation scenario. We show how a student model achieves state-of-the-art results on unsupervised domain adaptation from multiple sources on a standard sentiment analysis benchmark by taking into account the domain-specific expertise of multiple teachers and the similarities between their domains. When learning from a single teacher, using domain similarity to gauge trustworthiness is inadequate. To this end, we propose a simple metric that correlates well with the teacher's accuracy in the target domain. We demonstrate that incorporating high-confidence examples selected by this metric enables the student model to achieve state-of-the-art performance in the single-source scenario.

arXiv:1612.03928공식 코드 · Papers with Code

Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer

Attention plays a critical role in human visual experience. Furthermore, it has recently been demonstrated that attention can also play an important role in the context of applying artificial neural networks to a variety of tasks from fields such as computer vision and NLP. In this work we show that, by properly defining attention for convolutional neural networks, we can actually use this type of information in order to significantly improve the performance of a student CNN network by forcing it to mimic the attention maps of a powerful teacher network. To that end, we propose several novel methods of transferring attention, showing consistent improvement across a variety of datasets and convolutional neural network architectures. Code and models for our experiments are available at https://github.com/szagoruyko/attention-transfer

arXiv:1702.05796공식 코드 · Papers with Code

Collaborative Deep Reinforcement Learning

Besides independent learning, human learning process is highly improved by summarizing what has been learned, communicating it with peers, and subsequently fusing knowledge from different sources to assist the current learning goal. This collaborative learning procedure ensures that the knowledge is shared, continuously refined, and concluded from different perspectives to construct a more profound understanding. The idea of knowledge transfer has led to many advances in machine learning and data mining, but significant challenges remain, especially when it comes to reinforcement learning, heterogeneous model structures, and different learning tasks. Motivated by human collaborative learning, in this paper we propose a collaborative deep reinforcement learning (CDRL) framework that performs adaptive knowledge transfer among heterogeneous learning agents. Specifically, the proposed CDRL conducts a novel deep knowledge distillation method to address the heterogeneity among different learning tasks with a deep alignment network. Furthermore, we present an efficient collaborative Asynchronous Advantage Actor-Critic (cA3C) algorithm to incorporate deep knowledge distillation into the online training of agents, and demonstrate the effectiveness of the CDRL framework using extensive empirical evaluation on OpenAI gym.

arXiv:1703.07131

Knowledge distillation using unlabeled mismatched images

Current approaches for Knowledge Distillation (KD) either directly use training data or sample from the training data distribution. In this paper, we demonstrate effectiveness of 'mismatched' unlabeled stimulus to perform KD for image classification networks. For illustration, we consider scenarios where this is a complete absence of training data, or mismatched stimulus has to be used for augmenting a small amount of training data. We demonstrate that stimulus complexity is a key factor for distillation's good performance. Our examples include use of various datasets for stimulating MNIST and CIFAR teachers.

arXiv:1707.09423

Visual Relationship Detection with Internal and External Linguistic Knowledge Distillation

Understanding visual relationships involves identifying the subject, the object, and a predicate relating them. We leverage the strong correlations between the predicate and the (subj,obj) pair (both semantically and spatially) to predict the predicates conditioned on the subjects and the objects. Modeling the three entities jointly more accurately reflects their relationships, but complicates learning since the semantic space of visual relationships is huge and the training data is limited, especially for the long-tail relationships that have few instances. To overcome this, we use knowledge of linguistic statistics to regularize visual model learning. We obtain linguistic knowledge by mining from both training annotations (internal knowledge) and publicly available text, e.g., Wikipedia (external knowledge), computing the conditional probability distribution of a predicate given a (subj,obj) pair. Then, we distill the knowledge into a deep model to achieve better generalization. Our experimental results on the Visual Relationship Detection (VRD) and Visual Genome datasets suggest that with this linguistic knowledge distillation, our model outperforms the state-of-the-art methods significantly, especially when predicting unseen relationships (e.g., recall improved from 8.45% to 19.17% on VRD zero-shot testing set).

arXiv:1702.01802

Ensemble Distillation for Neural Machine Translation

Knowledge distillation describes a method for training a student network to perform better by learning from a stronger teacher network. Translating a sentence with an Neural Machine Translation (NMT) engine is time expensive and having a smaller model speeds up this process. We demonstrate how to transfer the translation quality of an ensemble and an oracle BLEU teacher network into a single NMT system. Further, we present translation improvements from a teacher network that has the same architecture and dimensions of the student network. As the training of the student model is still expensive, we introduce a data filtering method based on the knowledge of the teacher model that not only speeds up the training, but also leads to better translation quality. Our techniques need no code change and can be easily reproduced with any NMT architecture to speed up the decoding process.

arXiv:1604.00433

Adapting Models to Signal Degradation using Distillation

Model compression and knowledge distillation have been successfully applied for cross-architecture and cross-domain transfer learning. However, a key requirement is that training examples are in correspondence across the domains. We show that in many scenarios of practical importance such aligned data can be synthetically generated using computer graphics pipelines allowing domain adaptation through distillation. We apply this technique to learn models for recognizing low-resolution images using labeled high-resolution images, non-localized objects using labeled localized objects, line-drawings using labeled color images, etc. Experiments on various fine-grained recognition datasets demonstrate that the technique improves recognition performance on the low-quality data and beats strong baselines for domain adaptation. Finally, we present insights into workings of the technique through visualizations and relating it to existing literature.

arXiv:1709.02929

Model Distillation with Knowledge Transfer from Face Classification to Alignment and Verification

Knowledge distillation is a potential solution for model compression. The idea is to make a small student network imitate the target of a large teacher network, then the student network can be competitive to the teacher one. Most previous studies focus on model distillation in the classification task, where they propose different architects and initializations for the student network. However, only the classification task is not enough, and other related tasks such as regression and retrieval are barely considered. To solve the problem, in this paper, we take face recognition as a breaking point and propose model distillation with knowledge transfer from face classification to alignment and verification. By selecting appropriate initializations and targets in the knowledge transfer, the distillation can be easier in non-classification tasks. Experiments on the CelebA and CASIA-WebFace datasets demonstrate that the student network can be competitive to the teacher one in alignment and verification, and even surpasses the teacher network under specific compression rates. In addition, to achieve stronger knowledge transfer, we also use a common initialization trick to improve the distillation performance of classification. Evaluations on the CASIA-Webface and large-scale MS-Celeb-1M datasets show the effectiveness of this simple trick.

arXiv:1711.05852

Apprentice: Using Knowledge Distillation Techniques To Improve Low-Precision Network Accuracy

Deep learning networks have achieved state-of-the-art accuracies on computer vision workloads like image classification and object detection. The performant systems, however, typically involve big models with numerous parameters. Once trained, a challenging aspect for such top performing models is deployment on resource constrained inference systems - the models (often deep networks or wide networks or both) are compute and memory intensive. Low-precision numerics and model compression using knowledge distillation are popular techniques to lower both the compute requirements and memory footprint of these deployed models. In this paper, we study the combination of these two techniques and show that the performance of low-precision networks can be significantly improved by using knowledge distillation techniques. Our approach, Apprentice, achieves state-of-the-art accuracies using ternary precision and 4-bit precision for variants of ResNet architecture on ImageNet dataset. We present three schemes using which one can apply knowledge distillation techniques to various stages of the train-and-deploy pipeline.

arXiv:1711.07011

MicroExpNet: An Extremely Small and Fast Model For Expression Recognition From Face Images

This paper is aimed at creating extremely small and fast convolutional neural networks (CNN) for the problem of facial expression recognition (FER) from frontal face images. To this end, we employed the popular knowledge distillation (KD) method and identified two major shortcomings with its use: 1) a fine-grained grid search is needed for tuning the temperature hyperparameter and 2) to find the optimal size-accuracy balance, one needs to search for the final network size (or the compression rate). On the other hand, KD is proved to be useful for model compression for the FER problem, and we discovered that its effects gets more and more significant with the decreasing model size. In addition, we hypothesized that translation invariance achieved using max-pooling layers would not be useful for the FER problem as the expressions are sensitive to small, pixel-wise changes around the eye and the mouth. However, we have found an intriguing improvement on generalization when max-pooling is used. We conducted experiments on two widely-used FER datasets, CK+ and Oulu-CASIA. Our smallest model (MicroExpNet), obtained using knowledge distillation, is less than 1MB in size and works at 1851 frames per second on an Intel i7 CPU. Despite being less accurate than the state-of-the-art, MicroExpNet still provides significant insights for designing a microarchitecture for the FER problem.

arXiv:1711.07607

Knowledge Concentration: Learning 100K Object Classifiers in a Single CNN

Fine-grained image labels are desirable for many computer vision applications, such as visual search or mobile AI assistant. These applications rely on image classification models that can produce hundreds of thousands (e.g. 100K) of diversified fine-grained image labels on input images. However, training a network at this vocabulary scale is challenging, and suffers from intolerable large model size and slow training speed, which leads to unsatisfying classification performance. A straightforward solution would be training separate expert networks (specialists), with each specialist focusing on learning one specific vertical (e.g. cars, birds...). However, deploying dozens of expert networks in a practical system would significantly increase system complexity and inference latency, and consumes large amounts of computational resources. To address these challenges, we propose a Knowledge Concentration method, which effectively transfers the knowledge from dozens of specialists (multiple teacher networks) into one single model (one student network) to classify 100K object categories. There are three salient aspects in our method: (1) a multi-teacher single-student knowledge distillation framework; (2) a self-paced learning mechanism to allow the student to learn from different teachers at various paces; (3) structurally connected layers to expand the student network capacity with limited extra parameters. We validate our method on OpenImage and a newly collected dataset, Entity-Foto-Tree (EFT), with 100K categories, and show that the proposed model performs significantly better than the baseline generalist model.

arXiv:1710.07535

Data-Free Knowledge Distillation for Deep Neural Networks

Recent advances in model compression have provided procedures for compressing large neural networks to a fraction of their original size while retaining most if not all of their accuracy. However, all of these approaches rely on access to the original training set, which might not always be possible if the network to be compressed was trained on a very large dataset, or on a dataset whose release poses privacy or safety concerns as may be the case for biometrics tasks. We present a method for data-free knowledge distillation, which is able to compress deep neural networks trained on large-scale datasets to a fraction of their size leveraging only some extra metadata to be provided with a pretrained model release. We also explore different kinds of metadata that can be used with our method, and discuss tradeoffs involved in using each of them.

arXiv:1710.09282

A Survey of Model Compression and Acceleration for Deep Neural Networks

Deep neural networks (DNNs) have recently achieved great success in many visual recognition tasks. However, existing deep neural network models are computationally expensive and memory intensive, hindering their deployment in devices with low memory resources or in applications with strict latency requirements. Therefore, a natural thought is to perform model compression and acceleration in deep networks without significantly decreasing the model performance. During the past five years, tremendous progress has been made in this area. In this paper, we review the recent techniques for compacting and accelerating DNN models. In general, these techniques are divided into four categories: parameter pruning and quantization, low-rank factorization, transferred/compact convolutional filters, and knowledge distillation. Methods of parameter pruning and quantization are described first, after that the other techniques are introduced. For each category, we also provide insightful analysis about the performance, related applications, advantages, and drawbacks. Then we go through some very recent successful methods, for example, dynamic capacity networks and stochastic depths networks. After that, we survey the evaluation matrices, the main datasets used for evaluating the model performance, and recent benchmark efforts. Finally, we conclude this paper, discuss remaining the challenges and possible directions for future work.

arXiv:1706.02952

TIP: Typifying the Interpretability of Procedures

We provide a novel notion of what it means to be interpretable, looking past the usual association with human understanding. Our key insight is that interpretability is not an absolute concept and so we define it relative to a target model, which may or may not be a human. We define a framework that allows for comparing interpretable procedures by linking them to important practical aspects such as accuracy and robustness. We characterize many of the current state-of-the-art interpretable methods in our framework portraying its general applicability. Finally, principled interpretable strategies are proposed and empirically evaluated on synthetic data, as well as on the largest public olfaction dataset that was made recently available \cite{olfs}. We also experiment on MNIST with a simple target model and different oracle models of varying complexity. This leads to the insight that the improvement in the target model is not only a function of the oracle model's performance, but also its relative complexity with respect to the target model. Further experiments on CIFAR-10, a real manufacturing dataset and FICO dataset showcase the benefit of our methods over Knowledge Distillation when the target models are simple and the complex model is a neural network.

arXiv:1712.10136

Learning Deep and Compact Models for Gesture Recognition

We look at the problem of developing a compact and accurate model for gesture recognition from videos in a deep-learning framework. Towards this we propose a joint 3DCNN-LSTM model that is end-to-end trainable and is shown to be better suited to capture the dynamic information in actions. The solution achieves close to state-of-the-art accuracy on the ChaLearn dataset, with only half the model size. We also explore ways to derive a much more compact representation in a knowledge distillation framework followed by model compression. The final model is less than $1~MB$ in size, which is less than one hundredth of our initial model, with a drop of $7\%$ in accuracy, and is suitable for real-time gesture recognition on mobile devices.

arXiv:1801.05787

Faster gaze prediction with dense networks and Fisher pruning

Predicting human fixations from images has recently seen large improvements by leveraging deep representations which were pretrained for object recognition. However, as we show in this paper, these networks are highly overparameterized for the task of fixation prediction. We first present a simple yet principled greedy pruning method which we call Fisher pruning. Through a combination of knowledge distillation and Fisher pruning, we obtain much more runtime-efficient architectures for saliency prediction, achieving a 10x speedup for the same AUC performance as a state of the art network on the CAT2000 dataset. Speeding up single-image gaze prediction is important for many real-world applications, but it is also a crucial step in the development of video saliency models, where the amount of data to be processed is substantially larger.

arXiv:1801.10597

Model compression for faster structural separation of macromolecules captured by Cellular Electron Cryo-Tomography

Electron Cryo-Tomography (ECT) enables 3D visualization of macromolecule structure inside single cells. Macromolecule classification approaches based on convolutional neural networks (CNN) were developed to separate millions of macromolecules captured from ECT systematically. However, given the fast accumulation of ECT data, it will soon become necessary to use CNN models to efficiently and accurately separate substantially more macromolecules at the prediction stage, which requires additional computational costs. To speed up the prediction, we compress classification models into compact neural networks with little in accuracy for deployment. Specifically, we propose to perform model compression through knowledge distillation. Firstly, a complex teacher network is trained to generate soft labels with better classification feasibility followed by training of customized student networks with simple architectures using the soft label to compress model complexity. Our tests demonstrate that our compressed models significantly reduce the number of parameters and time cost while maintaining similar classification accuracy.