PyTorchKR Papers

최신 논문

총 1,324편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1808.06396

DeeSIL: Deep-Shallow Incremental Learning

Incremental Learning (IL) is an interesting AI problem when the algorithm is assumed to work on a budget. This is especially true when IL is modeled using a deep learning approach, where two com- plex challenges arise due to limited memory, which induces catastrophic forgetting and delays related to the retraining needed in order to incorpo- rate new classes. Here we introduce DeeSIL, an adaptation of a known transfer learning scheme that combines a fixed deep representation used as feature extractor and learning independent shallow classifiers to in- crease recognition capacity. This scheme tackles the two aforementioned challenges since it works well with a limited memory budget and each new concept can be added within a minute. Moreover, since no deep re- training is needed when the model is incremented, DeeSIL can integrate larger amounts of initial data that provide more transferable features. Performance is evaluated on ImageNet LSVRC 2012 against three state of the art algorithms. Results show that, at scale, DeeSIL performance is 23 and 33 points higher than the best baseline when using the same and more initial data respectively.

arXiv:1808.06281공식 코드 · Papers with Code

Incremental Learning in Person Re-Identification

Person Re-Identification is still a challenging task in Computer Vision due to a variety of reasons. On the other side, Incremental Learning is still an issue since deep learning models tend to face the problem of over catastrophic forgetting when trained on subsequent tasks. In this paper, we propose a model that can be used for multiple tasks in Person Re-Identification, provide state-of-the-art results on a variety of tasks and still achieve considerable accuracy subsequently. We evaluated our model on two datasets Market 1501 and Duke MTMC. Extensive experiments show that this method can achieve Incremental Learning in Person ReID efficiently as well as for other tasks in computer vision as well.

arXiv:1808.06048공식 코드 · Papers with Code

Distractor-aware Siamese Networks for Visual Object Tracking

Recently, Siamese networks have drawn great attention in visual tracking community because of their balanced accuracy and speed. However, features used in most Siamese tracking approaches can only discriminate foreground from the non-semantic backgrounds. The semantic backgrounds are always considered as distractors, which hinders the robustness of Siamese trackers. In this paper, we focus on learning distractor-aware Siamese networks for accurate and long-term tracking. To this end, features used in traditional Siamese trackers are analyzed at first. We observe that the imbalanced distribution of training data makes the learned features less discriminative. During the off-line training phase, an effective sampling strategy is introduced to control this distribution and make the model focus on the semantic distractors. During inference, a novel distractor-aware module is designed to perform incremental learning, which can effectively transfer the general embedding to the current video domain. In addition, we extend the proposed approach for long-term tracking by introducing a simple yet effective local-to-global search region strategy. Extensive experiments on benchmarks show that our approach significantly outperforms the state-of-the-arts, yielding 9.6% relative gain in VOT2016 dataset and 35.9% relative gain in UAV20L dataset. The proposed tracker can perform at 160 FPS on short-term benchmarks and 110 FPS on long-term benchmarks.

arXiv:1807.10211

Robust Tracking via Weighted Online Extreme Learning Machine

The tracking method based on the extreme learning machine (ELM) is efficient and effective. ELM randomly generates input weights and biases in the hidden layer, and then calculates and computes the output weights by reducing the iterative solution to the problem of linear equations. Therefore, ELM offers the satisfying classification performance and fast training time than other discriminative models in tracking. However, the original ELM method often suffers from the problem of the imbalanced classification distribution, which is caused by few target objects, leading to under-fitting and more background samples leading to over-fitting. Worse still, it reduces the robustness of tracking under special conditions including occlusion, illumination, etc. To address above problems, in this paper, we present a robust tracking algorithm. First, we introduce the local weight matrix that is the dynamic creation from the data distribution at the current frame in the original ELM so as to balance between the empirical and structure risk, and fully learn the target object to enhance the classification performance. Second, we improve it to the incremental learning method ensuring tracking real-time and efficient. Finally, the forgetting factor is used to strengthen the robustness for changing of the classification distribution with time. Meanwhile, we propose a novel optimized method to obtain the optimal sample as the target object, which avoids tracking drift resulting from noisy samples. Therefore, our tracking method can fully learn both of the target object and background information to enhance the tracking performance, and it is evaluated in 20 challenge image sequences with different attributes including illumination, occlusion, deformation, etc., which achieves better performance than several state-of-the-art methods in terms of effectiveness and robustness.

arXiv:1807.09536

End-to-End Incremental Learning

Although deep learning approaches have stood out in recent years due to their state-of-the-art results, they continue to suffer from catastrophic forgetting, a dramatic decrease in overall performance when training with new classes added incrementally. This is due to current neural network architectures requiring the entire dataset, consisting of all the samples from the old as well as the new classes, to update the model -a requirement that becomes easily unsustainable as the number of classes grows. We address this issue with our approach to learn deep neural networks incrementally, using new data and only a small exemplar set corresponding to samples from the old classes. This is based on a loss composed of a distillation measure to retain the knowledge acquired from the old classes, and a cross-entropy loss to learn the new classes. Our incremental training is achieved while keeping the entire framework end-to-end, i.e., learning the data representation and the classifier jointly, unlike recent methods with no such guarantees. We evaluate our method extensively on the CIFAR-100 and ImageNet (ILSVRC 2012) image classification datasets, and show state-of-the-art performance.

arXiv:1807.08942

Example Mining for Incremental Learning in Medical Imaging

Incremental Learning is well known machine learning approach wherein the weights of the learned model are dynamically and gradually updated to generalize on new unseen data without forgetting the existing knowledge. Incremental learning proves to be time as well as resource-efficient solution for deployment of deep learning algorithms in real world as the model can automatically and dynamically adapt to new data as and when annotated data becomes available. The development and deployment of Computer Aided Diagnosis (CAD) tools in medical domain is another scenario, where incremental learning becomes very crucial as collection and annotation of a comprehensive dataset spanning over multiple pathologies and imaging machines might take years. However, not much has so far been explored in this direction. In the current work, we propose a robust and efficient method for incremental learning in medical imaging domain. Our approach makes use of Hard Example Mining technique (which is commonly used as a solution to heavy class imbalance) to automatically select a subset of dataset to fine-tune the existing network weights such that it adapts to new data while retaining existing knowledge. We develop our approach for incremental learning of our already under test model for detecting dental caries. Further, we apply our approach to one publicly available dataset and demonstrate that our approach reaches the accuracy of training on entire dataset at once, while availing the benefits of incremental learning scenario.

arXiv:1807.03888공식 코드 · Papers with Code

A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks

Detecting test samples drawn sufficiently far away from the training distribution statistically or adversarially is a fundamental requirement for deploying a good classifier in many real-world machine learning applications. However, deep neural networks with the softmax classifier are known to produce highly overconfident posterior distributions even for such abnormal samples. In this paper, we propose a simple yet effective method for detecting any abnormal samples, which is applicable to any pre-trained softmax neural classifier. We obtain the class conditional Gaussian distributions with respect to (low- and upper-level) features of the deep models under Gaussian discriminant analysis, which result in a confidence score based on the Mahalanobis distance. While most prior methods have been evaluated for detecting either out-of-distribution or adversarial samples, but not both, the proposed method achieves the state-of-the-art performances for both cases in our experiments. Moreover, we found that our proposed method is more robust in harsh cases, e.g., when the training dataset has noisy labels or small number of samples. Finally, we show that the proposed method enjoys broader usage by applying it to class-incremental learning: whenever out-of-distribution samples are detected, our classification rule can incorporate new classes well without further training deep models.

arXiv:1807.02802공식 코드 · Papers with Code

Revisiting Distillation and Incremental Classifier Learning

One of the key differences between the learning mechanism of humans and Artificial Neural Networks (ANNs) is the ability of humans to learn one task at a time. ANNs, on the other hand, can only learn multiple tasks simultaneously. Any attempts at learning new tasks incrementally cause them to completely forget about previous tasks. This lack of ability to learn incrementally, called Catastrophic Forgetting, is considered a major hurdle in building a true AI system. In this paper, our goal is to isolate the truly effective existing ideas for incremental learning from those that only work under certain conditions. To this end, we first thoroughly analyze the current state of the art (iCaRL) method for incremental learning and demonstrate that the good performance of the system is not because of the reasons presented in the existing literature. We conclude that the success of iCaRL is primarily due to knowledge distillation and recognize a key limitation of knowledge distillation, i.e, it often leads to bias in classifiers. Finally, we propose a dynamic threshold moving algorithm that is able to successfully remove this bias. We demonstrate the effectiveness of our algorithm on CIFAR100 and MNIST datasets showing near-optimal results. Our implementation is available at https://github.com/Khurramjaved96/incremental-learning.

arXiv:1807.02799공식 코드 · Papers with Code

Distillation Techniques for Pseudo-rehearsal Based Incremental Learning

The ability to learn from incrementally arriving data is essential for any life-long learning system. However, standard deep neural networks forget the knowledge about the old tasks, a phenomenon called catastrophic forgetting, when trained on incrementally arriving data. We discuss the biases in current Generative Adversarial Networks (GAN) based approaches that learn the classifier by knowledge distillation from previously trained classifiers. These biases cause the trained classifier to perform poorly. We propose an approach to remove these biases by distilling knowledge from the classifier of AC-GAN. Experiments on MNIST and CIFAR10 show that this method is comparable to current state of the art rehearsal based approaches. The code for this paper is available at https://bit.ly/incremental-learning

arXiv:1806.09695

Person Re-Identification in Identity Regression Space

Most existing person re-identification (re-id) methods are unsuitable for real-world deployment due to two reasons: Unscalability to large population size, and Inadaptability over time. In this work, we present a unified solution to address both problems. Specifically, we propose to construct an Identity Regression Space (IRS) based on embedding different training person identities (classes) and formulate re-id as a regression problem solved by identity regression in the IRS. The IRS approach is characterised by a closed-form solution with high learning efficiency and an inherent incremental learning capability with human-in-the-loop. Extensive experiments on four benchmarking datasets(VIPeR, CUHK01, CUHK03 and Market-1501) show that the IRS model not only outperforms state-of-the-art re-id methods, but also is more scalable to large re-id population size by rapidly updating model and actively selecting informative samples with reduced human labelling effort.

arXiv:1806.09751공식 코드 · Papers with Code

A Practical Incremental Learning Framework For Sparse Entity Extraction

This work addresses challenges arising from extracting entities from textual data, including the high cost of data annotation, model accuracy, selecting appropriate evaluation criteria, and the overall quality of annotation. We present a framework that integrates Entity Set Expansion (ESE) and Active Learning (AL) to reduce the annotation cost of sparse data and provide an online evaluation method as feedback. This incremental and interactive learning framework allows for rapid annotation and subsequent extraction of sparse data while maintaining high accuracy. We evaluate our framework on three publicly available datasets and show that it drastically reduces the cost of sparse entity annotation by an average of 85% and 45% to reach 0.9 and 1.0 F-Scores respectively. Moreover, the method exhibited robust performance across all datasets.

arXiv:1806.08568

Continuous Learning in Single-Incremental-Task Scenarios

It was recently shown that architectural, regularization and rehearsal strategies can be used to train deep models sequentially on a number of disjoint tasks without forgetting previously acquired knowledge. However, these strategies are still unsatisfactory if the tasks are not disjoint but constitute a single incremental task (e.g., class-incremental learning). In this paper we point out the differences between multi-task and single-incremental-task scenarios and show that well-known approaches such as LWF, EWC and SI are not ideal for incremental task scenarios. A new approach, denoted as AR1, combining architectural and regularization strategies is then specifically proposed. AR1 overhead (in term of memory and computation) is very small thus making it suitable for online learning. When tested on CORe50 and iCIFAR-100, AR1 outperformed existing regularization strategies by a good margin.

arXiv:1304.3433

Induction, of and by Probability

This paper examines some methods and ideas underlying the author's successful probabilistic learning systems(PLS), which have proven uniquely effective and efficient in generalization learning or induction. While the emerging principles are generally applicable, this paper illustrates them in heuristic search, which demands noise management and incremental learning. In our approach, both task performance and learning are guided by probability. Probabilities are incrementally normalized and revised, and their errors are located and corrected.

arXiv:1405.3382

Active Mining of Parallel Video Streams

The practicality of a video surveillance system is adversely limited by the amount of queries that can be placed on human resources and their vigilance in response. To transcend this limitation, a major effort under way is to include software that (fully or at least semi) automatically mines video footage, reducing the burden imposed to the system. Herein, we propose a semi-supervised incremental learning framework for evolving visual streams in order to develop a robust and flexible track classification system. Our proposed method learns from consecutive batches by updating an ensemble in each time. It tries to strike a balance between performance of the system and amount of data which needs to be labelled. As no restriction is considered, the system can address many practical problems in an evolving multi-camera scenario, such as concept drift, class evolution and various length of video streams which have not been addressed before. Experiments were performed on synthetic as well as real-world visual data in non-stationary environments, showing high accuracy with fairly little human collaboration.

arXiv:1408.1487

Robust Feature Selection by Mutual Information Distributions

Mutual information is widely used in artificial intelligence, in a descriptive way, to measure the stochastic dependence of discrete random variables. In order to address questions such as the reliability of the empirical value, one must consider sample-to-population inferential approaches. This paper deals with the distribution of mutual information, as obtained in a Bayesian framework by a second-order Dirichlet prior distribution. The exact analytical expression for the mean and an analytical approximation of the variance are reported. Asymptotic approximations of the distribution are proposed. The results are applied to the problem of selecting features for incremental learning and classification of the naive Bayes classifier. A fast, newly defined method is shown to outperform the traditional approach based on empirical mutual information on a number of real data sets. Finally, a theoretical development is reported that allows one to efficiently extend the above methods to incomplete samples in an easy and effective way.

arXiv:1410.6776

Online and Stochastic Gradient Methods for Non-decomposable Loss Functions

Modern applications in sensitive domains such as biometrics and medicine frequently require the use of non-decomposable loss functions such as precision@k, F-measure etc. Compared to point loss functions such as hinge-loss, these offer much more fine grained control over prediction, but at the same time present novel challenges in terms of algorithm design and analysis. In this work we initiate a study of online learning techniques for such non-decomposable loss functions with an aim to enable incremental learning as well as design scalable solvers for batch problems. To this end, we propose an online learning framework for such loss functions. Our model enjoys several nice properties, chief amongst them being the existence of efficient online learning algorithms with sublinear regret and online to batch conversion bounds. Our model is a provable extension of existing online learning models for point loss functions. We instantiate two popular losses, prec@k and pAUC, in our model and prove sublinear regret bounds for both of them. Our proofs require a novel structural lemma over ranked lists which may be of independent interest. We then develop scalable stochastic gradient descent solvers for non-decomposable loss functions. We show that for a large family of loss functions satisfying a certain uniform convergence property (that includes prec@k, pAUC, and F-measure), our methods provably converge to the empirical risk minimizer. Such uniform convergence results were not known for these losses and we establish these using novel proof techniques. We then use extensive experimentation on real life and benchmark datasets to establish that our method can be orders of magnitude faster than a recently proposed cutting plane method.

arXiv:1402.5988

Incremental Learning of Event Definitions with Inductive Logic Programming

Event recognition systems rely on properly engineered knowledge bases of event definitions to infer occurrences of events in time. The manual development of such knowledge is a tedious and error-prone task, thus event-based applications may benefit from automated knowledge construction techniques, such as Inductive Logic Programming (ILP), which combines machine learning with the declarative and formal semantics of First-Order Logic. However, learning temporal logical formalisms, which are typically utilized by logic-based Event Recognition systems is a challenging task, which most ILP systems cannot fully undertake. In addition, event-based data is usually massive and collected at different times and under various circumstances. Ideally, systems that learn from temporal data should be able to operate in an incremental mode, that is, revise prior constructed knowledge in the face of new evidence. Most ILP systems are batch learners, in the sense that in order to account for new evidence they have no alternative but to forget past knowledge and learn from scratch. Given the increased inherent complexity of ILP and the volumes of real-life temporal data, this results to algorithms that scale poorly. In this work we present an incremental method for learning and revising event-based knowledge, in the form of Event Calculus programs. The proposed algorithm relies on abductive-inductive learning and comprises a scalable clause refinement methodology, based on a compressive summarization of clause coverage in a stream of examples. We present an empirical evaluation of our approach on real and synthetic data from activity recognition and city transport applications.

arXiv:1504.02870

Quick sensitivity analysis for incremental data modification and its application to leave-one-out CV in linear classification problems

We introduce a novel sensitivity analysis framework for large scale classification problems that can be used when a small number of instances are incrementally added or removed. For quickly updating the classifier in such a situation, incremental learning algorithms have been intensively studied in the literature. Although they are much more efficient than solving the optimization problem from scratch, their computational complexity yet depends on the entire training set size. It means that, if the original training set is large, completely solving an incremental learning problem might be still rather expensive. To circumvent this computational issue, we propose a novel framework that allows us to make an inference about the updated classifier without actually re-optimizing it. Specifically, the proposed framework can quickly provide a lower and an upper bounds of a quantity on the unknown updated classifier. The main advantage of the proposed framework is that the computational cost of computing these bounds depends only on the number of updated instances. This property is quite advantageous in a typical sensitivity analysis task where only a small number of instances are updated. In this paper we demonstrate that the proposed framework is applicable to various practical sensitivity analysis tasks, and the bounds provided by the framework are often sufficiently tight for making desired inferences.

arXiv:1506.04422

A Fast Incremental Gaussian Mixture Model

This work builds upon previous efforts in online incremental learning, namely the Incremental Gaussian Mixture Network (IGMN). The IGMN is capable of learning from data streams in a single-pass by improving its model after analyzing each data point and discarding it thereafter. Nevertheless, it suffers from the scalability point-of-view, due to its asymptotic time complexity of $\operatorname{O}\bigl(NKD^3\bigr)$ for $N$ data points, $K$ Gaussian components and $D$ dimensions, rendering it inadequate for high-dimensional data. In this paper, we manage to reduce this complexity to $\operatorname{O}\bigl(NKD^2\bigr)$ by deriving formulas for working directly with precision matrices instead of covariance matrices. The final result is a much faster and scalable algorithm which can be applied to high dimensional tasks. This is confirmed by applying the modified algorithm to high-dimensional classification datasets.

arXiv:1506.03229공식 코드 · Papers with Code

A cognitive neural architecture able to learn and communicate through natural language

Communicative interactions involve a kind of procedural knowledge that is used by the human brain for processing verbal and nonverbal inputs and for language production. Although considerable work has been done on modeling human language abilities, it has been difficult to bring them together to a comprehensive tabula rasa system compatible with current knowledge of how verbal information is processed in the brain. This work presents a cognitive system, entirely based on a large-scale neural architecture, which was developed to shed light on the procedural knowledge involved in language elaboration. The main component of this system is the central executive, which is a supervising system that coordinates the other components of the working memory. In our model, the central executive is a neural network that takes as input the neural activation states of the short-term memory and yields as output mental actions, which control the flow of information among the working memory components through neural gating mechanisms. The proposed system is capable of learning to communicate through natural language starting from tabula rasa, without any a priori knowledge of the structure of phrases, meaning of words, role of the different classes of words, only by interacting with a human through a text-based interface, using an open-ended incremental learning process. It is able to learn nouns, verbs, adjectives, pronouns and other word classes, and to use them in expressive language. The model was validated on a corpus of 1587 input sentences, based on literature on early language assessment, at the level of about 4-years old child, and produced 521 output sentences, expressing a broad range of language processing functionalities.