PyTorchKR Papers

최신 논문

총 1,324편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1906.01120공식 코드 · Papers with Code

An Adaptive Random Path Selection Approach for Incremental Learning

In a conventional supervised learning setting, a machine learning model has access to examples of all object classes that are desired to be recognized during the inference stage. This results in a fixed model that lacks the flexibility to adapt to new learning tasks. In practical settings, learning tasks often arrive in a sequence and the models must continually learn to increment their previously acquired knowledge. Existing incremental learning approaches fall well below the state-of-the-art cumulative models that use all training classes at once. In this paper, we propose a random path selection algorithm, called Adaptive RPS-Net, that progressively chooses optimal paths for the new tasks while encouraging parameter sharing between tasks. We introduce a new network capacity measure that enables us to automatically switch paths if the already used resources are saturated. Since the proposed path-reuse strategy ensures forward knowledge transfer, our approach is efficient and has considerably less computation overhead. As an added novelty, the proposed model integrates knowledge distillation and retrospection along with the path selection strategy to overcome catastrophic forgetting. In order to maintain an equilibrium between previous and newly acquired knowledge, we propose a simple controller to dynamically balance the model plasticity. Through extensive experiments, we demonstrate that the Adaptive RPS-Net method surpasses the state-of-the-art performance for incremental learning and by utilizing parallel computation this method can run in constant time with nearly the same efficiency as a conventional deep convolutional neural network.

arXiv:1906.00108

ActiveHARNet: Towards On-Device Deep Bayesian Active Learning for Human Activity Recognition

Various health-care applications such as assisted living, fall detection etc., require modeling of user behavior through Human Activity Recognition (HAR). HAR using mobile- and wearable-based deep learning algorithms have been on the rise owing to the advancements in pervasive computing. However, there are two other challenges that need to be addressed: first, the deep learning model should support on-device incremental training (model updation) from real-time incoming data points to learn user behavior over time, while also being resource-friendly; second, a suitable ground truthing technique (like Active Learning) should help establish labels on-the-fly while also selecting only the most informative data points to query from an oracle. Hence, in this paper, we propose ActiveHARNet, a resource-efficient deep ensembled model which supports on-device Incremental Learning and inference, with capabilities to represent model uncertainties through approximations in Bayesian Neural Networks using dropout. This is combined with suitable acquisition functions for active learning. Empirical results on two publicly available wrist-worn HAR and fall detection datasets indicate that ActiveHARNet achieves considerable efficiency boost during inference across different users, with a substantially low number of acquired pool points (at least 60% reduction) during incremental learning on both datasets experimented with various acquisition functions, thus demonstrating deployment and Incremental Learning feasibility.

arXiv:1905.13260

Large Scale Incremental Learning

Modern machine learning suffers from catastrophic forgetting when learning new classes incrementally. The performance dramatically degrades due to the missing data of old classes. Incremental learning methods have been proposed to retain the knowledge acquired from the old classes, by using knowledge distilling and keeping a few exemplars from the old classes. However, these methods struggle to scale up to a large number of classes. We believe this is because of the combination of two factors: (a) the data imbalance between the old and new classes, and (b) the increasing number of visually similar classes. Distinguishing between an increasing number of visually similar classes is particularly challenging, when the training data is unbalanced. We propose a simple and effective method to address this data imbalance issue. We found that the last fully connected layer has a strong bias towards the new classes, and this bias can be corrected by a linear model. With two bias parameters, our method performs remarkably well on two large datasets: ImageNet (1000 classes) and MS-Celeb-1M (10000 classes), outperforming the state-of-the-art algorithms by 11.1% and 13.2% respectively.

arXiv:1905.12628

Personalizing human activity recognition models using incremental learning

In this study, the aim is to personalize inertial sensor data-based human activity recognition models using incremental learning. At first, the recognition is based on user-independent model. However, when personal streaming data becomes available, the incremental learning-based recognition model can be updated, and therefore personalized, based on the data without user-interruption. The used incremental learning algorithm is Learn++ which is an ensemble method that can use any classifier as a base classifier. In fact, study compares three different base classifiers: linear discriminant analysis (LDA), quadratic discriminant analysis (QDA) and classification and regression tree (CART). Experiments are based on publicly open data set and they show that already a small personal training data set can improve the classification accuracy. Improvement using LDA as base classifier is 4.6 percentage units, using QDA 2.0 percentage units, and 2.3 percentage units using CART. However, if the user-independent model used in the first phase of the recognition process is not accurate enough, personalization cannot improve recognition accuracy.

arXiv:1905.12181

Leveraging Semantics for Incremental Learning in Multi-Relational Embeddings

Service robots benefit from encoding information in semantically meaningful ways to enable more robust task execution. Prior work has shown multi-relational embeddings can encode semantic knowledge graphs to promote generalizability and scalability, but only within a batched learning paradigm. We present Incremental Semantic Initialization (ISI), an incremental learning approach that enables novel semantic concepts to be initialized in the embedding in relation to previously learned embeddings of semantically similar concepts. We evaluate ISI on mined AI2Thor and MatterPort3D datasets; our experiments show that on average ISI improves immediate query performance by 41.4%. Additionally, ISI methods on average reduced the number of epochs required to approach model convergence by 78.2%.

arXiv:1905.11775

Importance of user inputs while using incremental learning to personalize human activity recognition models

In this study, importance of user inputs is studied in the context of personalizing human activity recognition models using incremental learning. Inertial sensor data from three body positions are used, and the classification is based on Learn++ ensemble method. Three different approaches to update models are compared: non-supervised, semi-supervised and supervised. Non-supervised approach relies fully on predicted labels, supervised fully on user labeled data, and the proposed method for semi-supervised learning, is a combination of these two. In fact, our experiments show that by relying on predicted labels with high confidence, and asking the user to label only uncertain observations (from 12% to 26% of the observations depending on the used base classifier), almost as low error rates can be achieved as by using supervised approach. In fact, the difference was less than 2%-units. Moreover, unlike non-supervised approach, semi-supervised approach does not suffer from drastic concept drift, and thus, the error rate of the non-supervised approach is over 5%-units higher than using semi-supervised approach.

arXiv:1905.10952

Incremental Learning Using a Grow-and-Prune Paradigm with Efficient Neural Networks

Deep neural networks (DNNs) have become a widely deployed model for numerous machine learning applications. However, their fixed architecture, substantial training cost, and significant model redundancy make it difficult to efficiently update them to accommodate previously unseen data. To solve these problems, we propose an incremental learning framework based on a grow-and-prune neural network synthesis paradigm. When new data arrive, the neural network first grows new connections based on the gradients to increase the network capacity to accommodate new data. Then, the framework iteratively prunes away connections based on the magnitude of weights to enhance network compactness, and hence recover efficiency. Finally, the model rests at a lightweight DNN that is both ready for inference and suitable for future grow-and-prune updates. The proposed framework improves accuracy, shrinks network size, and significantly reduces the additional training cost for incoming data compared to conventional approaches, such as training from scratch and network fine-tuning. For the LeNet-300-100 and LeNet-5 neural network architectures derived for the MNIST dataset, the framework reduces training cost by up to 64% (63%) and 67% (63%) compared to training from scratch (network fine-tuning), respectively. For the ResNet-18 architecture derived for the ImageNet dataset and DeepSpeech2 for the AN4 dataset, the corresponding training cost reductions against training from scratch (network fine-tunning) are 64% (60%) and 67% (62%), respectively. Our derived models contain fewer network parameters but achieve higher accuracy relative to conventional baselines.

arXiv:1905.08101

A comprehensive, application-oriented study of catastrophic forgetting in DNNs

We present a large-scale empirical study of catastrophic forgetting (CF) in modern Deep Neural Network (DNN) models that perform sequential (or: incremental) learning. A new experimental protocol is proposed that enforces typical constraints encountered in application scenarios. As the investigation is empirical, we evaluate CF behavior on the hitherto largest number of visual classification datasets, from each of which we construct a representative number of Sequential Learning Tasks (SLTs) in close alignment to previous works on CF. Our results clearly indicate that there is no model that avoids CF for all investigated datasets and SLTs under application conditions. We conclude with a discussion of potential solutions and workarounds to CF, notably for the EWC and IMM models.

arXiv:1905.08077

Catastrophic forgetting: still a problem for DNNs

We investigate the performance of DNNs when trained on class-incremental visual problems consisting of initial training, followed by retraining with added visual classes. Catastrophic forgetting (CF) behavior is measured using a new evaluation procedure that aims at an application-oriented view of incremental learning. In particular, it imposes that model selection must be performed on the initial dataset alone, as well as demanding that retraining control be performed only using the retraining dataset, as initial dataset is usually too large to be kept. Experiments are conducted on class-incremental problems derived from MNIST, using a variety of different DNN models, some of them recently proposed to avoid catastrophic forgetting. When comparing our new evaluation procedure to previous approaches for assessing CF, we find their findings are completely negated, and that none of the tested methods can avoid CF in all experiments. This stresses the importance of a realistic empirical measurement procedure for catastrophic forgetting, and the need for further research in incremental learning for DNNs.

arXiv:1905.07835

Label Mapping Neural Networks with Response Consolidation for Class Incremental Learning

Class incremental learning refers to a special multi-class classification task, in which the number of classes is not fixed but is increasing with the continual arrival of new data. Existing researches mainly focused on solving catastrophic forgetting problem in class incremental learning. To this end, however, these models still require the old classes cached in the auxiliary data structure or models, which is inefficient in space or time. In this paper, it is the first time to discuss the difficulty without support of old classes in class incremental learning, which is called as softmax suppression problem. To address these challenges, we develop a new model named Label Mapping with Response Consolidation (LMRC), which need not access the old classes anymore. We propose the Label Mapping algorithm combined with the multi-head neural network for mitigating the softmax suppression problem, and propose the Response Consolidation method to overcome the catastrophic forgetting problem. Experimental results on the benchmark datasets show that our proposed method achieves much better performance compared to the related methods in different scenarios.

arXiv:1905.04010

A fast online cascaded regression algorithm for face alignment

Traditional face alignment based on machine learning usually tracks the localizations of facial landmarks employing a static model trained offline where all of the training data is available in advance. When new training samples arrive, the static model must be retrained from scratch, which is excessively time-consuming and memory-consuming. In many real-time applications, the training data is obtained one by one or batch by batch. It results in that the static model limits its performance on sequential images with extensive variations. Therefore, the most critical and challenging aspect in this field is dynamically updating the tracker's models to enhance predictive and generalization capabilities continuously. In order to address this question, we develop a fast and accurate online learning algorithm for face alignment. Particularly, we incorporate on-line sequential extreme learning machine into a parallel cascaded regression framework, coined incremental cascade regression(ICR). To the best of our knowledge, this is the first incremental cascaded framework with the non-linear regressor. One main advantage of ICR is that the tracker model can be fast updated in an incremental way without the entire retraining process when a new input is incoming. Experimental results demonstrate that the proposed ICR is more accurate and efficient on still or sequential images compared with the recent state-of-the-art cascade approaches. Furthermore, the incremental learning proposed in this paper can update the trained model in real time.

arXiv:1905.00921

Continuous Learning for Large-scale Personalized Domain Classification

Domain classification is the task of mapping spoken language utterances to one of the natural language understanding domains in intelligent personal digital assistants (IPDAs). This is a major component in mainstream IPDAs in industry. Apart from official domains, thousands of third-party domains are also created by external developers to enhance the capability of IPDAs. As more domains are developed rapidly, the question of how to continuously accommodate the new domains still remains challenging. Moreover, existing continual learning approaches do not address the problem of incorporating personalized information dynamically for better domain classification. In this paper, we propose CoNDA, a neural network based approach for domain classification that supports incremental learning of new classes. Empirical evaluation shows that CoNDA achieves high accuracy and outperforms baselines by a large margin on both incrementally added new domains and existing domains.

arXiv:1904.12118

Incremental personalized E-mail spam filter using novel TFDCR feature selection with dynamic feature update

Communication through e-mails remains to be highly formalized, conventional and indispensable method for the exchange of information over the Internet. An ever-increasing ratio and adversary nature of spam e-mails have posed a great many challenges such as uneven class distribution, unequal error cost, frequent change of content and personalized context-sensitive discrimination. In this research, we propose a novel and distinctive approach to develop an incremental personalized e-mail spam filter. The proposed work is described using three significant contributions. First, we applied a novel term frequency difference and category ratio based feature selection function TFDCR to select the most discriminating features irrespective of the number of samples in each class. Second, an incremental learning model is used which enables the classifier to update the discriminant function dynamically. Third, a heuristic function called selectionRankWeight is introduced to upgrade the existing feature set that determines new features carrying strong discriminating ability from an incoming set of e-mails. Three public e-mail datasets possessing different characteristics are used to evaluate the filter performance. Experiments are conducted to compare the feature selection efficiency of TFDCR and to observe the filter performance under both the batch and the incremental learning mode. The results demonstrate the superiority of TFDCR as the most effective f eature selection function. The incremental learning model incorporating dynamic feature update function overcomes the problem of drifting concepts. The proposed filter validates its efficiency and feasibility by substantially improving the classification accuracy and reducing the false positive error of misclassifying legitimate e-mail as spam.

arXiv:1904.07734공식 코드 · Papers with Code

Three scenarios for continual learning

Standard artificial neural networks suffer from the well-known issue of catastrophic forgetting, making continual or lifelong learning difficult for machine learning. In recent years, numerous methods have been proposed for continual learning, but due to differences in evaluation protocols it is difficult to directly compare their performance. To enable more structured comparisons, we describe three continual learning scenarios based on whether at test time task identity is provided and--in case it is not--whether it must be inferred. Any sequence of well-defined tasks can be performed according to each scenario. Using the split and permuted MNIST task protocols, for each scenario we carry out an extensive comparison of recently proposed continual learning methods. We demonstrate substantial differences between the three scenarios in terms of difficulty and in terms of how efficient different methods are. In particular, when task identity must be inferred (i.e., class incremental learning), we find that regularization-based approaches (e.g., elastic weight consolidation) fail and that replaying representations of previous experiences seems required for solving this scenario.

arXiv:1904.06345

Incremental multi-domain learning with network latent tensor factorization

The prominence of deep learning, large amount of annotated data and increasingly powerful hardware made it possible to reach remarkable performance for supervised classification tasks, in many cases saturating the training sets. However the resulting models are specialized to a single very specific task and domain. Adapting the learned classification to new domains is a hard problem due to at least three reasons: (1) the new domains and the tasks might be drastically different; (2) there might be very limited amount of annotated data on the new domain and (3) full training of a new model for each new task is prohibitive in terms of computation and memory, due to the sheer number of parameters of deep CNNs. In this paper, we present a method to learn new-domains and tasks incrementally, building on prior knowledge from already learned tasks and without catastrophic forgetting. We do so by jointly parametrizing weights across layers using low-rank Tucker structure. The core is task agnostic while a set of task specific factors are learnt on each new domain. We show that leveraging tensor structure enables better performance than simply using matrix operations. Joint tensor modelling also naturally leverages correlations across different layers. Compared with previous methods which have focused on adapting each layer separately, our approach results in more compact representations for each new task/domain. We apply the proposed method to the 10 datasets of the Visual Decathlon Challenge and show that our method offers on average about 7.5x reduction in number of parameters and competitive performance in terms of both classification accuracy and Decathlon score.

arXiv:1904.03423공식 코드 · Papers with Code

FILDNE: A Framework for Incremental Learning of Dynamic Networks Embeddings

Representation learning on graphs has emerged as a powerful mechanism to automate feature vector generation for downstream machine learning tasks. The advances in representation on graphs have centered on both homogeneous and heterogeneous graphs, where the latter presenting the challenges associated with multi-typed nodes and/or edges. In this paper, we consider the additional challenge of evolving graphs. We ask the question of whether the advances in representation learning for static graphs can be leveraged for dynamic graphs and how? It is important to be able to incorporate those advances to maximize the utility and generalization of methods. To that end, we propose the Framework for Incremental Learning of Dynamic Networks Embedding (FILDNE), which can utilize any existing static representation learning method for learning node embeddings, while keeping the computational costs low. FILDNE integrates the feature vectors computed using the standard methods over different timesteps into a single representation by developing a convex combination function and alignment mechanism. Experimental results on several downstream tasks, over seven real-world data sets, show that FILDNE is able to reduce memory and computational time costs while providing competitive quality measure gains with respect to the contemporary methods for representation learning on dynamic graphs.

arXiv:1904.02420

Transfer Learning with Sparse Associative Memories

In this paper, we introduce a novel layer designed to be used as the output of pre-trained neural networks in the context of classification. Based on Associative Memories, this layer can help design Deep Neural Networks which support incremental learning and that can be (partially) trained in real time on embedded devices. Experiments on the ImageNet dataset and other different domain specific datasets show that it is possible to design more flexible and faster-to-train Neural Networks at the cost of a slight decrease in accuracy.

arXiv:1904.01769

M2KD: Multi-model and Multi-level Knowledge Distillation for Incremental Learning

Incremental learning targets at achieving good performance on new categories without forgetting old ones. Knowledge distillation has been shown critical in preserving the performance on old classes. Conventional methods, however, sequentially distill knowledge only from the last model, leading to performance degradation on the old classes in later incremental learning steps. In this paper, we propose a multi-model and multi-level knowledge distillation strategy. Instead of sequentially distilling knowledge only from the last model, we directly leverage all previous model snapshots. In addition, we incorporate an auxiliary distillation to further preserve knowledge encoded at the intermediate feature levels. To make the model more memory efficient, we adapt mask based pruning to reconstruct all previous models with a small memory footprint. Experiments on standard incremental learning benchmarks show that our method preserves the knowledge on old classes better and improves the overall performance over standard distillation techniques.

arXiv:1904.00781

RILOD: Near Real-Time Incremental Learning for Object Detection at the Edge

Object detection models shipped with camera-equipped edge devices cannot cover the objects of interest for every user. Therefore, the incremental learning capability is a critical feature for a robust and personalized object detection system that many applications would rely on. In this paper, we present an efficient yet practical system, RILOD, to incrementally train an existing object detection model such that it can detect new object classes without losing its capability to detect old classes. The key component of RILOD is a novel incremental learning algorithm that trains end-to-end for one-stage deep object detection models only using training data of new object classes. Specifically to avoid catastrophic forgetting, the algorithm distills three types of knowledge from the old model to mimic the old model's behavior on object classification, bounding box regression and feature extraction. In addition, since the training data for the new classes may not be available, a real-time dataset construction pipeline is designed to collect training images on-the-fly and automatically label the images with both category and bounding box annotations. We have implemented RILOD under both edge-cloud and edge-only setups. Experiment results show that the proposed system can learn to detect a new object class in just a few minutes, including both dataset construction and model training. In comparison, traditional fine-tuning based method may take a few hours for training, and in most cases would also need a tedious and costly manual dataset labeling step.

arXiv:1903.12648공식 코드 · Papers with Code

Overcoming Catastrophic Forgetting with Unlabeled Data in the Wild

Lifelong learning with deep neural networks is well-known to suffer from catastrophic forgetting: the performance on previous tasks drastically degrades when learning a new task. To alleviate this effect, we propose to leverage a large stream of unlabeled data easily obtainable in the wild. In particular, we design a novel class-incremental learning scheme with (a) a new distillation loss, termed global distillation, (b) a learning strategy to avoid overfitting to the most recent task, and (c) a confidence-based sampling method to effectively leverage unlabeled external data. Our experimental results on various datasets, including CIFAR and ImageNet, demonstrate the superiority of the proposed methods over prior methods, particularly when a stream of unlabeled data is accessible: our method shows up to 15.8% higher accuracy and 46.5% less forgetting compared to the state-of-the-art method. The code is available at https://github.com/kibok90/iccv2019-inc.