In this paper, we propose an end-to-end Mandarin tone classification method from continuous speech utterances utilizing both the spectrogram and the short-term context information as the input. Both spectrograms and context segment features are used to train the tone classifier. We first divide the spectrogram frames into syllable segments using force alignment results produced by an ASR model. Then we extract the short-term segment features to capture the context information across multiple syllables. Feeding both the spectrogram and the short-term context segment features into an end-to-end model could significantly improve the performance. Experiments are performed on a large-scale open-source Mandarin speech dataset to evaluate the proposed method. Results show that this method improves the classification accuracy from 79.5% to 92.6% on the AISHELL3 database.
A challenging part of dynamic probabilistic risk assessment for nuclear power plants is the need for large amounts of temporal simulations given various initiating events and branching conditions from which representative feature extraction becomes complicated for subsequent applications. Artificial Intelligence techniques have been shown to be powerful tools in time-dependent sequential data processing to automatically extract and yield complex features from large data. An advanced temporal neural network referred to as the Transformer is used within a supervised learning fashion to model the time-dependent NPP simulation data and to infer whether a given sequence of events leads to core damage or not. The training and testing datasets for the Transformer are obtained by running 10,000 RELAP5-3D NPP blackout simulations with the list of variables obtained from the RAVEN software. Each simulation is classified as "OK" or "CORE DAMAGE" based on the consequence. The results show that the Transformer can learn the characteristics of the sequential data and yield promising performance with approximately 99% classification accuracy on the testing dataset.
In satellite image analysis, distributional mismatch between the training and test data may arise due to several reasons, including unseen classes in the test data and differences in the geographic area. Deep learning based models may behave in unexpected manner when subjected to test data that has such distributional shifts from the training data, also called out-of-distribution (OOD) examples. Predictive uncertainly analysis is an emerging research topic which has not been explored much in context of satellite image analysis. Towards this, we adopt a Dirichlet Prior Network based model to quantify distributional uncertainty of deep learning models for remote sensing. The approach seeks to maximize the representation gap between the in-domain and OOD examples for a better identification of unknown examples at test time. Experimental results on three exemplary test scenarios show the efficacy of the model in satellite image analysis.
The current state-of-the-art model HiAGM for hierarchical text classification has two limitations. First, it correlates each text sample with all labels in the dataset which contains irrelevant information. Second, it does not consider any statistical constraint on the label representations learned by the structure encoder, while constraints for representation learning are proved to be helpful in previous work. In this paper, we propose HTCInfoMax to address these issues by introducing information maximization which includes two modules: text-label mutual information maximization and label prior matching. The first module can model the interaction between each text sample and its ground truth labels explicitly which filters out irrelevant information. The second one encourages the structure encoder to learn better representations with desired characteristics for all labels which can better handle label imbalance in hierarchical text classification. Experimental results on two benchmark datasets demonstrate the effectiveness of the proposed HTCInfoMax.
We study the problem of automated mechanism design with partial verification, where each type can (mis)report only a restricted set of types (rather than any other type), induced by the principal's limited verification power. We prove hardness results when the revelation principle does not necessarily hold, as well as when types have even minimally different preferences. In light of these hardness results, we focus on truthful mechanisms in the setting where all types share the same preference over outcomes, which is motivated by applications in, e.g., strategic classification. We present a number of algorithmic and structural results, including an efficient algorithm for finding optimal deterministic truthful mechanisms, which also implies a faster algorithm for finding optimal randomized truthful mechanisms via a characterization based on convexity. We then consider a more general setting, where the principal's cost is a function of the combination of outcomes assigned to each type. In particular, we focus on the case where the cost function is submodular, and give generalizations of essentially all our results in the classical setting where the cost function is additive. Our results provide a relatively complete picture for automated mechanism design with partial verification.
Contrastive Learning has emerged as a powerful representation learning method and facilitates various downstream tasks especially when supervised data is limited. How to construct efficient contrastive samples through data augmentation is key to its success. Unlike vision tasks, the data augmentation method for contrastive learning has not been investigated sufficiently in language tasks. In this paper, we propose a novel approach to construct contrastive samples for language tasks using text summarization. We use these samples for supervised contrastive learning to gain better text representations which greatly benefit text classification tasks with limited annotations. To further improve the method, we mix up samples from different classes and add an extra regularization, named Mixsum, in addition to the cross-entropy-loss. Experiments on real-world text classification datasets (Amazon-5, Yelp-5, AG News, and IMDb) demonstrate the effectiveness of the proposed contrastive learning framework with summarization-based data augmentation and Mixsum regularization.
An increasing number of emerging applications in data science and engineering are based on multidimensional and structurally rich data. The irregularities, however, of high-dimensional data often compromise the effectiveness of standard machine learning algorithms. We hereby propose the Rank-R Feedforward Neural Network (FNN), a tensor-based nonlinear learning model that imposes Canonical/Polyadic decomposition on its parameters, thereby offering two core advantages compared to typical machine learning methods. First, it handles inputs as multilinear arrays, bypassing the need for vectorization, and can thus fully exploit the structural information along every data dimension. Moreover, the number of the model's trainable parameters is substantially reduced, making it very efficient for small sample setting problems. We establish the universal approximation and learnability properties of Rank-R FNN, and we validate its performance on real-world hyperspectral datasets. Experimental evaluations show that Rank-R FNN is a computationally inexpensive alternative of ordinary FNN that achieves state-of-the-art performance on higher-order tensor data.
In multi-class classification tasks, like human activity recognition, it is often assumed that classes are separable. In real applications, this assumption becomes strong and generates inconsistencies. Besides, the most commonly used approach is to learn classes one-by-one against the others. This computational simplification principle introduces strong inductive biases on the learned theories. In fact, the natural connections among some classes, and not others, deserve to be taken into account. In this paper, we show that the organization of overlapping classes (multiple inheritances) into hierarchies considerably improves classification performances. This is particularly true in the case of activity recognition tasks featured in the SHL dataset. After theoretically showing the exponential complexity of possible class hierarchies, we propose an approach based on transfer affinity among the classes to determine an optimal hierarchy for the learning process. Extensive experiments show improved performances and a reduction in the number of examples needed to learn.
Contemporary Artificial Intelligence technologies allow for the employment of Computer Vision to discern good crops from bad, providing a step in the pipeline of selecting healthy fruit from undesirable fruit, such as those which are mouldy or gangrenous. State-of-the-art works in the field report high accuracy results on small datasets (<1000 images), which are not representative of the population regarding real-world usage. The goals of this study are to further enable real-world usage by improving generalisation with data augmentation as well as to reduce overfitting and energy usage through model pruning. In this work, we suggest a machine learning pipeline that combines the ideas of fine-tuning, transfer learning, and generative model-based training data augmentation towards improving fruit quality image classification. A linear network topology search is performed to tune a VGG16 lemon quality classification model using a publicly-available dataset of 2690 images. We find that appending a 4096 neuron fully connected layer to the convolutional layers leads to an image classification accuracy of 83.77%. We then train a Conditional Generative Adversarial Network on the training data for 2000 epochs, and it learns to generate relatively realistic images. Grad-CAM analysis of the model trained on real photographs shows that the synthetic images can exhibit classifiable characteristics such as shape, mould, and gangrene. A higher image classification accuracy of 88.75% is then attained by augmenting the training with synthetic images, arguing that Conditional Generative Adversarial Networks have the ability to produce new data to alleviate issues of data scarcity. Finally, model pruning is performed via polynomial decay, where we find that the Conditional GAN-augmented classification network can retain 81.16% classification accuracy when compressed to 50% of its original size.
Pulmonary optical endomicroscopy (POE) is an imaging technology in real time. It allows to examine pulmonary alveoli at a microscopic level. Acquired in clinical settings, a POE image sequence can have as much as 25% of the sequence being uninformative frames (i.e. pure-noise and motion artefacts). For future data analysis, these uninformative frames must be first removed from the sequence. Therefore, the objective of our work is to develop an automatic detection method of uninformative images in endomicroscopy images. We propose to take the detection problem as a classification one. Considering advantages of deep learning methods, a classifier based on CNN (Convolutional Neural Network) is designed with a new loss function based on Havrda-Charvat entropy which is a parametrical generalization of the Shannon entropy. We propose to use this formula to get a better hold on all sorts of data since it provides a model more stable than the Shannon entropy. Our method is tested on one POE dataset including 2947 distinct images, is showing better results than using Shannon entropy and behaves better with regard to the problem of overfitting. Keywords: Deep Learning, CNN, Shannon entropy, Havrda-Charvat entropy, Pulmonary optical endomicroscopy.
Most classification models can be considered as the process of matching templates. However, when intra-class uncertainty/variability is not considered, especially for datasets containing unbalanced classes, this may lead to classification errors. To address this issue, we propose a loss function with intra-class uncertainty following Gaussian distribution. Specifically, in our framework, the features extracted by deep networks of each class are characterized by independent Gaussian distribution. The parameters of distribution are learned with a likelihood regularization along with other network parameters. The means of the Gaussian play a similar role as the center anchor in existing methods, and the variance describes the uncertainty of different classes. In addition, similar to the inter-class margin in traditional loss functions, we introduce a margin to intra-class uncertainty to make each cluster more compact and reduce the imbalance of feature distribution from different categories. Based on MNIST, CIFAR, ImageNet, and Long-tailed CIFAR analyses, the proposed approach shows improved classification performance, through learning a better class representation.
We propose a two-stage Convolutional Neural Network (CNN) based classification framework for detecting COVID-19 and Community-Acquired Pneumonia (CAP) using the chest Computed Tomography (CT) scan images. In the first stage, an infection - COVID-19 or CAP, is detected using a pre-trained DenseNet architecture. Then, in the second stage, a fine-grained three-way classification is done using EfficientNet architecture. The proposed COVID+CAP-CNN framework achieved a slice-level classification accuracy of over 94% at identifying COVID-19 and CAP. Further, the proposed framework has the potential to be an initial screening tool for differential diagnosis of COVID-19 and CAP, achieving a validation accuracy of over 89.3% at the finer three-way COVID-19, CAP, and healthy classification. Within the IEEE ICASSP 2021 Signal Processing Grand Challenge (SPGC) on COVID-19 Diagnosis, our proposed two-stage classification framework achieved an overall accuracy of 90% and sensitivity of .857, .9, and .942 at distinguishing COVID-19, CAP, and normal individuals respectively, to rank first in the evaluation. Code and model weights are available at https://github.com/shubhamchaudhary2015/ct_covid19_cap_cnn
Zero-shot learning, the task of learning to recognize new classes not seen during training, has received considerable attention in the case of 2D image classification. However, despite the increasing ubiquity of 3D sensors, the corresponding 3D point cloud classification problem has not been meaningfully explored and introduces new challenges. In this paper, we identify some of the challenges and apply 2D Zero-Shot Learning (ZSL) methods in the 3D domain to analyze the performance of existing models. Then, we propose a novel approach to address the issues specific to 3D ZSL. We first present an inductive ZSL process and then extend it to the transductive ZSL and Generalized ZSL (GZSL) settings for 3D point cloud classification. To this end, a novel loss function is developed that simultaneously aligns seen semantics with point cloud features and takes advantage of unlabeled test data to address some known issues (e.g., the problems of domain adaptation, hubness, and data bias). While designed for the particularities of 3D point cloud classification, the method is shown to also be applicable to the more common use-case of 2D image classification. An extensive set of experiments is carried out, establishing state-of-the-art for ZSL and GZSL on synthetic (ModelNet40, ModelNet10, McGill) and real (ScanObjectNN) 3D point cloud datasets.
Kubelka-Munk (K-M) theory has been successfully used to estimate pigment concentrations in the pigment mixtures of modern paintings in spectral imagery. In this study the single-constant K-M theory has been utilized for the classification of green pigments in the Selden Map of China, a navigational map of the South China Sea likely created in the early seventeenth century. Hyperspectral data of the map was collected at the Bodleian Library, University of Oxford, and can be used to estimate the pigment diversity, and spatial distribution, within the map. This work seeks to assess the utility of analyzing the data in the K/S space from Kubelka-Munk theory, as opposed to the traditional reflectance domain. We estimate the dimensionality of the data and extract endmembers in the reflectance domain. Then we perform linear unmixing to estimate abundances in the K/S space, and following Bai, et al. (2017), we perform a classification in the abundance space. Finally, due to the lack of ground truth labels, the classification accuracy was estimated by computing the mean spectrum of each class as the representative signature of that class, and calculating the root mean squared error with all the pixels in that class to create a spatial representation of the error. This highlights both the magnitude of, and any spatial pattern in, the errors, indicating if a particular pigment is not well modeled in this approach.
Continual learning has become increasingly important as it enables NLP models to constantly learn and gain knowledge over time. Previous continual learning methods are mainly designed to preserve knowledge from previous tasks, without much emphasis on how to well generalize models to new tasks. In this work, we propose an information disentanglement based regularization method for continual learning on text classification. Our proposed method first disentangles text hidden spaces into representations that are generic to all tasks and representations specific to each individual task, and further regularizes these representations differently to better constrain the knowledge required to generalize. We also introduce two simple auxiliary tasks: next sentence prediction and task-id prediction, for learning better generic and specific representation spaces. Experiments conducted on large-scale benchmarks demonstrate the effectiveness of our method in continual text classification tasks with various sequences and lengths over state-of-the-art baselines. We have publicly released our code at https://github.com/GT-SALT/IDBR.
Chest radiography is the most common radiographic examination performed in daily clinical practice for the detection of various heart and lung abnormalities. The large amount of data to be read and reported, with more than 100 studies per day for a single radiologist, poses a challenge in consistently maintaining high interpretation accuracy. The introduction of large-scale public datasets has led to a series of novel systems for automated abnormality classification. However, the labels of these datasets were obtained using natural language processed medical reports, yielding a large degree of label noise that can impact the performance. In this study, we propose novel training strategies that handle label noise from such suboptimal data. Prior label probabilities were measured on a subset of training data re-read by 4 board-certified radiologists and were used during training to increase the robustness of the training model to the label noise. Furthermore, we exploit the high comorbidity of abnormalities observed in chest radiography and incorporate this information to further reduce the impact of label noise. Additionally, anatomical knowledge is incorporated by training the system to predict lung and heart segmentation, as well as spatial knowledge labels. To deal with multiple datasets and images derived from various scanners that apply different post-processing techniques, we introduce a novel image normalization strategy. Experiments were performed on an extensive collection of 297,541 chest radiographs from 86,876 patients, leading to a state-of-the-art performance level for 17 abnormalities from 2 datasets. With an average AUC score of 0.880 across all abnormalities, our proposed training strategies can be used to significantly improve performance scores.
Patients with Frontotemporal Dementia (FTD) have impaired cognitive abilities, executive and behavioral traits, loss of language ability, and decreased memory capabilities. Based on the distinct patterns of cortical atrophy and symptoms, the FTD spectrum primarily includes three variants: behavioral variant FTD (bvFTD), non-fluent variant primary progressive aphasia (nfvPPA), and semantic variant primary progressive aphasia (svPPA). The purpose of this study is to classify MRI images of every single subject into one of the spectrums of the FTD in a hierarchical order by applying data-driven techniques of Artificial Intelligence (AI) on cortical thickness data. This data is computed by FreeSurfer software. We used the Smallest Univalue Segment Assimilating Nucleus (SUSAN) technique to minimize the noise in cortical thickness data. Specifically, we took 204 subjects from the frontotemporal lobar degeneration neuroimaging initiative (NIFTD) database to validate this approach, and each subject was diagnosed in one of the diagnostic categories (bvFTD, svPPA, nfvPPA and cognitively normal). Our proposed automated classification model yielded classification accuracy of 86.5, 76, and 72.7 with support vector machine (SVM), linear discriminant analysis (LDA), and Naive Bayes methods, respectively, in 10-fold cross-validation analysis, which is a significant improvement on a traditional single multi-class model with an accuracy of 82.7, 73.4, and 69.2.
Lipschitz constrained networks have gathered considerable attention in the deep learning community, with usages ranging from Wasserstein distance estimation to the training of certifiably robust classifiers. However they remain commonly considered as less accurate, and their properties in learning are still not fully understood. In this paper we clarify the matter: when it comes to classification 1-Lipschitz neural networks enjoy several advantages over their unconstrained counterpart. First, we show that these networks are as accurate as classical ones, and can fit arbitrarily difficult boundaries. Then, relying on a robustness metric that reflects operational needs we characterize the most robust classifier: the WGAN discriminator. Next, we show that 1-Lipschitz neural networks generalize well under milder assumptions. Finally, we show that hyper-parameters of the loss are crucial for controlling the accuracy-robustness trade-off. We conclude that they exhibit appealing properties to pave the way toward provably accurate, and provably robust neural networks.
Emotion recognition in conversations is an important step in various virtual chat bots which require opinion-based feedback, like in social media threads, online support and many more applications. Current Emotion recognition in conversations models face issues like (a) loss of contextual information in between two dialogues of a conversation, (b) failure to give appropriate importance to significant tokens in each utterance and (c) inability to pass on the emotional information from previous utterances.The proposed model of Advanced Contextual Feature Extraction (AdCOFE) addresses these issues by performing unique feature extraction using knowledge graphs, sentiment lexicons and phrases of natural language at all levels (word and position embedding) of the utterances. Experiments on the Emotion recognition in conversations dataset show that AdCOFE is beneficial in capturing emotions in conversations.
Speech based depression classification has gained immense popularity over the recent years. However, most of the classification studies have focused on binary classification to distinguish depressed subjects from non-depressed subjects. In this paper, we formulate the depression classification task as a severity level classification problem to provide more granularity to the classification outcomes. We use articulatory coordination features (ACFs) developed to capture the changes of neuromotor coordination that happens as a result of psychomotor slowing, a necessary feature of Major Depressive Disorder. The ACFs derived from the vocal tract variables (TVs) are used to train a dilated Convolutional Neural Network based depression classification model to obtain segment-level predictions. Then, we propose a Recurrent Neural Network based approach to obtain session-level predictions from segment-level predictions. We show that strengths of the segment-wise classifier are amplified when a session-wise classifier is trained on embeddings obtained from it. The model trained on ACFs derived from TVs show relative improvement of 27.47% in Unweighted Average Recall (UAR) at the session-level classification task, compared to the ACFs derived from Mel Frequency Cepstral Coefficients (MFCCs).