This work presents a novel fundamental algorithm for for defining and training Neural Networks in Quantum Information based on time evolution and the Hamiltonian. Classical Neural Network algorithms (ANN) are computationally expensive. For example, in image classification, representing an image pixel by pixel using classical information requires an enormous amount of computational memory resources. Hence, exploring methods to represent images in a different paradigm of information is important. Quantum Neural Networks (QNNs) have been explored for over 20 years. The current forefront work based on Variational Quantum Circuits is specifically defined for the Continuous Variable (CV) Model of quantum computers. In this work, a model is proposed which is defined at a more fundamental level and hence can be inherited by any variants of quantum computing models. This work also presents a quantum backpropagation algorithm to train our QNN model and validate this algorithm on the MNIST dataset on a quantum computer simulation.
Existing deep neural networks, say for image classification, have been shown to be vulnerable to adversarial images that can cause a DNN misclassification, without any perceptible change to an image. In this work, we propose shock absorbing robust features such as binarization, e.g., rounding, and group extraction, e.g., color or shape, to augment the classification pipeline, resulting in more robust classifiers. Experimentally, we show that augmenting ML models with these techniques leads to improved overall robustness on adversarial inputs as well as significant improvements in training time. On the MNIST dataset, we achieved 14x speedup in training time to obtain 90% adversarial accuracy com-pared to the state-of-the-art adversarial training method of Madry et al., as well as retained higher adversarial accuracy over a broader range of attacks. We also find robustness improvements on traffic sign classification using robust feature augmentation. Finally, we give theoretical insights for why one can expect robust feature augmentation to reduce adversarial input space
We report a series of robust empirical observations, demonstrating that deep Neural Networks learn the examples in both the training and test sets in a similar order. This phenomenon is observed in all the commonly used benchmarks we evaluated, including many image classification benchmarks, and one text classification benchmark. While this phenomenon is strongest for models of the same architecture, it also crosses architectural boundaries -- models of different architectures start by learning the same examples, after which the more powerful model may continue to learn additional examples. We further show that this pattern of results reflects the interplay between the way neural networks learn benchmark datasets. Thus, when fixing the architecture, we show synthetic datasets where this pattern ceases to exist. When fixing the dataset, we show that other learning paradigms may learn the data in a different order. We hypothesize that our results reflect how neural networks discover structure in natural datasets.
Explaining a deep learning model can help users understand its behavior and allow researchers to discern its shortcomings. Recent work has primarily focused on explaining models for tasks like image classification or visual question answering. In this paper, we introduce Salient Attributes for Network Explanation (SANE) to explain image similarity models, where a model's output is a score measuring the similarity of two inputs rather than a classification score. In this task, an explanation depends on both of the input images, so standard methods do not apply. Our SANE explanations pairs a saliency map identifying important image regions with an attribute that best explains the match. We find that our explanations provide additional information not typically captured by saliency maps alone, and can also improve performance on the classic task of attribute recognition. Our approach's ability to generalize is demonstrated on two datasets from diverse domains, Polyvore Outfits and Animals with Attributes 2. Code available at: https://github.com/VisionLearningGroup/SANE
Activation functions play an important role in training artificial neural networks. The majority of currently used activation functions are deterministic in nature, with their fixed input-output relationship. In this work, we propose a novel probabilistic activation function, called ProbAct. ProbAct is decomposed into a mean and variance and the output value is sampled from the formed distribution, making ProbAct a stochastic activation function. The values of mean and variances can be fixed using known functions or trained for each element. In the trainable ProbAct, the mean and the variance of the activation distribution is trained within the back-propagation framework alongside other parameters. We show that the stochastic perturbation induced through ProbAct acts as a viable generalization technique for feature augmentation. In our experiments, we compare ProbAct with well-known activation functions on classification tasks on different modalities: Images(CIFAR-10, CIFAR-100, and STL-10) and Text (Large Movie Review). We show that ProbAct increases the classification accuracy by +2-3% compared to ReLU or other conventional activation functions on both original datasets and when datasets are reduced to 50% and 25% of the original size. Finally, we show that ProbAct learns an ensemble of models by itself that can be used to estimate the uncertainties associated with the prediction and provides robustness to noisy inputs.
Mixup~\cite{zhang2017mixup} is a recently proposed method for training deep neural networks where additional samples are generated during training by convexly combining random pairs of images and their associated labels. While simple to implement, it has been shown to be a surprisingly effective method of data augmentation for image classification: DNNs trained with mixup show noticeable gains in classification performance on a number of image classification benchmarks. In this work, we discuss a hitherto untouched aspect of mixup training -- the calibration and predictive uncertainty of models trained with mixup. We find that DNNs trained with mixup are significantly better calibrated -- i.e., the predicted softmax scores are much better indicators of the actual likelihood of a correct prediction -- than DNNs trained in the regular fashion. We conduct experiments on a number of image classification architectures and datasets -- including large-scale datasets like ImageNet -- and find this to be the case. Additionally, we find that merely mixing features does not result in the same calibration benefit and that the label smoothing in mixup training plays a significant role in improving calibration. Finally, we also observe that mixup-trained DNNs are less prone to over-confident predictions on out-of-distribution and random-noise data. We conclude that the typical overconfidence seen in neural networks, even on in-distribution data is likely a consequence of training with hard labels, suggesting that mixup be employed for classification tasks where predictive uncertainty is a significant concern.
Real-world large-scale datasets usually contain noisy labels and are imbalanced. Therefore, we propose derivative manipulation (DM), a novel and general example weighting approach for training robust deep models under these adverse conditions. DM has two main merits. First, loss function and example weighting are two common techniques in robust learning. In gradient-based optimisation, the role of a loss function is to provide the gradient for back-propagation to update a model, so that the derivative magnitude of an example defines how much impact it has, namely its weight. By DM, we connect the design of loss function and example weighting together. Second, although designing a loss function sometimes has the same effect, we need to care whether a loss is differentiable, and derive its derivative to understand its example weighting scheme. They make the design complicated. Instead, DM is more flexible and straightforward by directly modifying the derivative. Concretely, DM modifies a derivative magnitude function, including transformation and normalisation, after which we term it an emphasis density function, which expresses a weighting scheme. Accordingly, diverse weighting schemes are derived from common probability density functions, including those of well-known robust losses, e.g., MAE and GCE. We conduct extensive experiments demonstrating the effectiveness of DM on both vision and language tasks.
We formalise the widespread idea of interpreting neural network decisions as an explicit optimisation problem in a rate-distortion framework. A set of input features is deemed relevant for a classification decision if the expected classifier score remains nearly constant when randomising the remaining features. We discuss the computational complexity of finding small sets of relevant features and show that the problem is complete for $\mathsf{NP}^\mathsf{PP}$, an important class of computational problems frequently arising in AI tasks. Furthermore, we show that it even remains $\mathsf{NP}$-hard to only approximate the optimal solution to within any non-trivial approximation factor. Finally, we consider a continuous problem relaxation and develop a heuristic solution strategy based on assumed density filtering for deep ReLU neural networks. We present numerical experiments for two image classification data sets where we outperform established methods in particular for sparse explanations of neural network decisions.
Convolutional neural networks (CNNs) have achieved remarkable performance in various fields, particularly in the domain of computer vision. However, why this architecture works well remains to be a mystery. In this work we move a small step toward understanding the success of CNNs by investigating the learning dynamics of a two-layer nonlinear convolutional neural network over some specific data distributions. Rather than the typical Gaussian assumption for input data distribution, we consider a more realistic setting that each data point (e.g. image) contains a specific pattern determining its class label. Within this setting, we both theoretically and empirically show that some convolutional filters will learn the key patterns in data and the norm of these filters will dominate during the training process with stochastic gradient descent. And with any high probability, when the number of iterations is sufficiently large, the CNN model could obtain 100% accuracy over the considered data distributions. Our experiments demonstrate that for practical image classification tasks our findings still hold to some extent.
Recently, Capsule Networks (CapsNets) have shown improved performance compared to the traditional Convolutional Neural Networks (CNNs), by encoding and preserving spatial relationships between the detected features in a better way. This is achieved through the so-called Capsules (i.e., groups of neurons) that encode both the instantiation probability and the spatial information. However, one of the major hurdles in the wide adoption of CapsNets is their gigantic training time, which is primarily due to the relatively higher complexity of their new constituting elements that are different from CNNs. In this paper, we implement different optimizations in the training loop of the CapsNets, and investigate how these optimizations affect their training speed and the accuracy. Towards this, we propose a novel framework FasTrCaps that integrates multiple lightweight optimizations and a novel learning rate policy called WarmAdaBatch (that jointly performs warm restarts and adaptive batch size), and steers them in an appropriate way to provide high training-loop speedup at minimal accuracy loss. We also propose weight sharing for capsule layers. The goal is to reduce the hardware requirements of CapsNets by removing unused/redundant connections and capsules, while keeping high accuracy through tests of different learning rate policies and batch sizes. We demonstrate that one of the solutions generated by the FasTrCaps framework can achieve 58.6% reduction in the training time, while preserving the accuracy (even 0.12% accuracy improvement for the MNIST dataset), compared to the CapsNet by Google Brain. The Pareto-optimal solutions generated by FasTrCaps can be leveraged to realize trade-offs between training time and achieved accuracy. We have open-sourced our framework on https://github.com/Alexei95/FasTrCaps.
Dropout is a simple but efficient regularization technique for achieving better generalization of deep neural networks (DNNs); hence it is widely used in tasks based on DNNs. During training, dropout randomly discards a portion of the neurons to avoid overfitting. This paper presents an enhanced dropout technique, which we call multi-sample dropout, for both accelerating training and improving generalization over the original dropout. The original dropout creates a randomly selected subset (called a dropout sample) from the input in each training iteration while the multi-sample dropout creates multiple dropout samples. The loss is calculated for each sample, and then the sample losses are averaged to obtain the final loss. This technique can be easily implemented by duplicating a part of the network after the dropout layer while sharing the weights among the duplicated fully connected layers. Experimental results using image classification tasks including ImageNet, CIFAR-10, and CIFAR-100 showed that multi-sample dropout accelerates training. Moreover, the networks trained using multi-sample dropout achieved lower error rates compared to networks trained with the original dropout. The additional computation cost due to the duplicated operations is not significant for deep convolutional networks because most of the computation time is consumed in the convolution layers before the dropout layer, which are not duplicated.
Convolutional neural networks (CNNs) have obtained astounding successes for important pattern recognition tasks, but they suffer from high computational complexity and the lack of interpretability. The recent Tsetlin Machine (TM) attempts to address this lack by using easy-to-interpret conjunctive clauses in propositional logic to solve complex pattern recognition problems. The TM provides competitive accuracy in several benchmarks, while keeping the important property of interpretability. It further facilitates hardware-near implementation since inputs, patterns, and outputs are expressed as bits, while recognition and learning rely on straightforward bit manipulation. In this paper, we exploit the TM paradigm by introducing the Convolutional Tsetlin Machine (CTM), as an interpretable alternative to CNNs. Whereas the TM categorizes an image by employing each clause once to the whole image, the CTM uses each clause as a convolution filter. That is, a clause is evaluated multiple times, once per image patch taking part in the convolution. To make the clauses location-aware, each patch is further augmented with its coordinates within the image. The output of a convolution clause is obtained simply by ORing the outcome of evaluating the clause on each patch. In the learning phase of the TM, clauses that evaluate to 1 are contrasted against the input. For the CTM, we instead contrast against one of the patches, randomly selected among the patches that made the clause evaluate to 1. Accordingly, the standard Type I and Type II feedback of the classic TM can be employed directly, without further modification. The CTM obtains a peak test accuracy of 99.4% on MNIST, 96.31% on Kuzushiji-MNIST, 91.5% on Fashion-MNIST, and 100.0% on the 2D Noisy XOR Problem, which is competitive with results reported for simple 4-layer CNNs, BinaryConnect, Logistic Circuits and an FPGA-accelerated Binary CNN.
The Convolutional Neural Networks (CNNs) generate the feature representation of complex objects by collecting hierarchical and different parts of semantic sub-features. These sub-features can usually be distributed in grouped form in the feature vector of each layer, representing various semantic entities. However, the activation of these sub-features is often spatially affected by similar patterns and noisy backgrounds, resulting in erroneous localization and identification. We propose a Spatial Group-wise Enhance (SGE) module that can adjust the importance of each sub-feature by generating an attention factor for each spatial location in each semantic group, so that every individual group can autonomously enhance its learnt expression and suppress possible noise. The attention factors are only guided by the similarities between the global and local feature descriptors inside each group, thus the design of SGE module is extremely lightweight with \emph{almost no extra parameters and calculations}. Despite being trained with only category supervisions, the SGE component is extremely effective in highlighting multiple active areas with various high-order semantics (such as the dog's eyes, nose, etc.). When integrated with popular CNN backbones, SGE can significantly boost the performance of image recognition tasks. Specifically, based on ResNet50 backbones, SGE achieves 1.2\% Top-1 accuracy improvement on the ImageNet benchmark and 1.0$\sim$2.0\% AP gain on the COCO benchmark across a wide range of detectors (Faster/Mask/Cascade RCNN and RetinaNet). Codes and pretrained models are available at https://github.com/implus/PytorchInsight.
Deep neural networks have achieved a great success in solving many machine learning and computer vision problems. The main contribution of this paper is to develop a deep network based on Tucker tensor decomposition, and analyze its expressive power. It is shown that the expressiveness of Tucker network is more powerful than that of shallow network. In general, it is required to use an exponential number of nodes in a shallow network in order to represent a Tucker network. Experimental results are also given to compare the performance of the proposed Tucker network with hierarchical tensor network and shallow network, and demonstrate the usefulness of Tucker network in image classification problems.
Deep neural networks have been shown to perform well in many classical machine learning problems, especially in image classification tasks. However, researchers have found that neural networks can be easily fooled, and they are surprisingly sensitive to small perturbations imperceptible to humans. Carefully crafted input images (adversarial examples) can force a well-trained neural network to provide arbitrary outputs. Including adversarial examples during training is a popular defense mechanism against adversarial attacks. In this paper we propose a new defensive mechanism under the generative adversarial network (GAN) framework. We model the adversarial noise using a generative network, trained jointly with a classification discriminative network as a minimax game. We show empirically that our adversarial network approach works well against black box attacks, with performance on par with state-of-art methods such as ensemble adversarial training and adversarial training with projected gradient descent.
Recently, Convolutional Neural Networks (CNNs) have shown unprecedented success in the field of computer vision, especially on challenging image classification tasks by relying on a universal approach, i.e., training a deep model on a massive dataset of supervised examples. While unlabeled data are often an abundant resource, collecting a large set of labeled data, on the other hand, are very expensive, which often require considerable human efforts. One way to ease out this is to effectively select and label highly informative instances from a pool of unlabeled data (i.e., active learning). This paper proposed a new method of batch-mode active learning, Dual Active Sampling(DAS), which is based on a simple assumption, if two deep neural networks (DNNs) of the same structure and trained on the same dataset give significantly different output for a given sample, then that particular sample should be picked for additional training. While other state of the art methods in this field usually require intensive computational power or relying on a complicated structure, DAS is simpler to implement and, managed to get improved results on Cifar-10 with preferable computational time compared to the core-set method.
In this paper, we propose an unified hyperspectral image classification method which takes three-dimensional hyperspectral data cube as an input and produces a classification map. In the proposed method, a deep neural network which uses spectral and spatial information together with residual connections, and pixel affinity network based segmentation-aware superpixels are used together. In the architecture, segmentation-aware superpixels run on the initial classification map of deep residual network, and apply majority voting on obtained results. Experimental results show that our propoped method yields state-of-the-art results in two benchmark datasets. Moreover, we also show that the segmentation-aware superpixels have great contribution to the success of hyperspectral image classification methods in cases where training data is insufficient.
Automatic analysis of scanned historical documents comprises a wide range of image analysis tasks, which are often challenging for machine learning due to a lack of human-annotated learning samples. With the advent of deep neural networks, a promising way to cope with the lack of training data is to pre-train models on images from a different domain and then fine-tune them on historical documents. In the current research, a typical example of such cross-domain transfer learning is the use of neural networks that have been pre-trained on the ImageNet database for object recognition. It remains a mostly open question whether or not this pre-training helps to analyse historical documents, which have fundamentally different image properties when compared with ImageNet. In this paper, we present a comprehensive empirical survey on the effect of ImageNet pre-training for diverse historical document analysis tasks, including character recognition, style classification, manuscript dating, semantic segmentation, and content-based retrieval. While we obtain mixed results for semantic segmentation at pixel-level, we observe a clear trend across different network architectures that ImageNet pre-training has a positive effect on classification as well as content-based retrieval.
In recent studies, several asymptotic upper bounds on generalization errors on deep neural networks (DNNs) are theoretically derived. These bounds are functions of several norms of weights of the DNNs, such as the Frobenius and spectral norms, and they are computed for weights grouped according to either input and output channels of the DNNs. In this work, we conjecture that if we can impose multiple constraints on weights of DNNs to upper bound the norms of the weights, and train the DNNs with these weights, then we can attain empirical generalization errors closer to the derived theoretical bounds, and improve accuracy of the DNNs. To this end, we pose two problems. First, we aim to obtain weights whose different norms are all upper bounded by a constant number, e.g. 1.0. To achieve these bounds, we propose a two-stage renormalization procedure; (i) normalization of weights according to different norms used in the bounds, and (ii) reparameterization of the normalized weights to set a constant and finite upper bound of their norms. In the second problem, we consider training DNNs with these renormalized weights. To this end, we first propose a strategy to construct joint spaces (manifolds) of weights according to different constraints in DNNs. Next, we propose a fine-grained SGD algorithm (FG-SGD) for optimization on the weight manifolds to train DNNs with assurance of convergence to minima. Experimental results show that image classification accuracy of baseline DNNs can be boosted using FG-SGD on collections of manifolds identified by multiple constraints.
Mixup, a recent proposed data augmentation method through linearly interpolating inputs and modeling targets of random samples, has demonstrated its capability of significantly improving the predictive accuracy of the state-of-the-art networks for image classification. However, how this technique can be applied to and what is its effectiveness on natural language processing (NLP) tasks have not been investigated. In this paper, we propose two strategies for the adaption of Mixup on sentence classification: one performs interpolation on word embeddings and another on sentence embeddings. We conduct experiments to evaluate our methods using several benchmark datasets. Our studies show that such interpolation strategies serve as an effective, domain independent data augmentation approach for sentence classification, and can result in significant accuracy improvement for both CNN and LSTM models.