Learning from 3D point-cloud data has rapidly gained momentum, motivated by the success of deep learning on images and the increased availability of 3D~data. In this paper, we aim to construct anisotropic convolution layers that work directly on the surface derived from a point cloud. This is challenging because of the lack of a global coordinate system for tangential directions on surfaces. We introduce DeltaConv, a convolution layer that combines geometric operators from vector calculus to enable the construction of anisotropic filters on point clouds. Because these operators are defined on scalar- and vector-fields, we separate the network into a scalar- and a vector-stream, which are connected by the operators. The vector stream enables the network to explicitly represent, evaluate, and process directional information. Our convolutions are robust and simple to implement and match or improve on state-of-the-art approaches on several benchmarks, while also speeding up training and inference.
Sinhala language is being the widely used language in Sri Lanka. With the advancement of internet usage in Sri Lanka, an incredible amount of Sinhala text data is being added to the internet. In order to manage, analyze and make decisions from the available text data, it requires text classification. Being a low resource and morphologically rich language requires higher expertise and a considerable amount of budget and time to develop an effective task-specific text classifier. This research aims to develop a domain or dataset agnostic and automated solution to improve the quality and address current research gaps of text classification in Sinhala. Based on the solution, a high-level development framework and a user interface are developed. In addition, we perform a cross-domain evaluation with multiple datasets to evaluate the effectiveness of the solution. The proposed framework achieved state-of-the-art results for the Sinhala text classification.
In this paper we present a novel mathematical optimization-based methodology to construct tree-shaped classification rules for multiclass instances. Our approach consists of building Classification Trees in which, except for the leaf nodes, the labels are temporarily left out and grouped into two classes by means of a SVM separating hyperplane. We provide a Mixed Integer Non Linear Programming formulation for the problem and report the results of an extended battery of computational experiments to assess the performance of our proposal with respect to other benchmarking classification methods.
Mitigating the energy requirements of artificial intelligence requires novel physical substrates for computation. Phononic metamaterials have a vanishingly low power dissipation and hence are a prime candidate for green, always-on computers. However, their use in machine learning applications has not been explored due to the complexity of their design process: Current phononic metamaterials are restricted to simple geometries (e.g. periodic, tapered), and hence do not possess sufficient expressivity to encode machine learning tasks. We design and fabricate a non-periodic phononic metamaterial, directly from data samples, that can distinguish between pairs of spoken words in the presence of a simple readout nonlinearity; hence demonstrating that phononic metamaterials are a viable avenue towards zero-power smart devices.
Deep Neural Networks (DNNs) have performed admirably in classification tasks. However, the characterization of their classification uncertainties, required for certain applications, has been lacking. In this work, we investigate the issue by assessing DNNs' ability to estimate conditional probabilities and propose a framework for systematic uncertainty characterization. Denoting the input sample as x and the category as y, the classification task of assigning a category y to a given input x can be reduced to the task of estimating the conditional probabilities p(y|x), as approximated by the DNN at its last layer using the softmax function. Since softmax yields a vector whose elements all fall in the interval (0, 1) and sum to 1, it suggests a probabilistic interpretation to the DNN's outcome. Using synthetic and real-world datasets, we look into the impact of various factors, e.g., probability density f(x) and inter-categorical sparsity, on the precision of DNNs' estimations of p(y|x), and find that the likelihood probability density and the inter-categorical sparsity have greater impacts than the prior probability to DNNs' classification uncertainty.
In this paper classification of mental task-root Brain-Computer Interfaces (BCI) is being investigated, as those are a dominant area of investigations in BCI and are of utmost interest as these systems can be augmented life of people having severe disabilities. The BCI model's performance is primarily dependent on the size of the feature vector, which is obtained through multiple channels. In the case of mental task classification, the availability of training samples to features are minimal. Very often, feature selection is used to increase the ratio for the mental task classification by getting rid of irrelevant and superfluous features. This paper proposes an approach to select relevant and non-redundant spectral features for the mental task classification. This can be done by using four very known multivariate feature selection methods viz, Bhattacharya's Distance, Ratio of Scatter Matrices, Linear Regression and Minimum Redundancy & Maximum Relevance. This work also deals with a comparative analysis of multivariate and univariate feature selection for mental task classification. After applying the above-stated method, the findings demonstrate substantial improvements in the performance of the learning model for mental task classification. Moreover, the efficacy of the proposed approach is endorsed by carrying out a robust ranking algorithm and Friedman's statistical test for finding the best combinations and comparing different combinations of power spectral density and feature selection methods.
We consider the problem of multi-class classification, where a stream of adversarially chosen queries arrive and must be assigned a label online. Unlike traditional bounds which seek to minimize the misclassification rate, we minimize the total distance from each query to the region corresponding to its correct label. When the true labels are determined via a nearest neighbor partition -- i.e. the label of a point is given by which of $k$ centers it is closest to in Euclidean distance -- we show that one can achieve a loss that is independent of the total number of queries. We complement this result by showing that learning general convex sets requires an almost linear loss per query. Our results build off of regret guarantees for the geometric problem of contextual search. In addition, we develop a novel reduction technique from multiclass classification to binary classification which may be of independent interest.
This paper considers the learning of Boolean rules in disjunctive normal form (DNF, OR-of-ANDs, equivalent to decision rule sets) as an interpretable model for classification. An integer program is formulated to optimally trade classification accuracy for rule simplicity. We also consider the fairness setting and extend the formulation to include explicit constraints on two different measures of classification parity: equality of opportunity and equalized odds. Column generation (CG) is used to efficiently search over an exponential number of candidate rules without the need for heuristic rule mining. To handle large data sets, we propose an approximate CG algorithm using randomization. Compared to three recently proposed alternatives, the CG algorithm dominates the accuracy-simplicity trade-off in 8 out of 16 data sets. When maximized for accuracy, CG is competitive with rule learners designed for this purpose, sometimes finding significantly simpler solutions that are no less accurate. Compared to other fair and interpretable classifiers, our method is able to find rule sets that meet stricter notions of fairness with a modest trade-off in accuracy.
In this paper, we investigate the impact of some of the commonly used settings for (a) preprocessing face images, and (b) classification and training, on Action Unit (AU) detection performance and complexity. We use in our investigation a large-scale dataset, consisting of ~55K videos collected in the wild for participants watching commercial ads. The preprocessing settings include scaling the face to a fixed resolution, changing the color information (RGB to gray-scale), aligning the face, and cropping AU regions, while the classification and training settings include the kind of classifier (multi-label vs. binary) and the amount of data used for training models. To the best of our knowledge, no work had investigated the effect of those settings on AU detection. In our analysis we use CNNs as our baseline classification model.
Despite being one of the most spoken languages in the world ($6^{th}$ based on population), research regarding Bengali handwritten grapheme (smallest functional unit of a writing system) classification has not been explored widely compared to other prominent languages. Moreover, the large number of combinations of graphemes in the Bengali language makes this classification task very challenging. With an effort to contribute to this research problem, we participate in a Kaggle competition \cite{kaggle_link} where the challenge is to separately classify three constituent elements of a Bengali grapheme in the image: grapheme root, vowel diacritics, and consonant diacritics. We explore the performances of some existing neural network models such as Multi-Layer Perceptron (MLP) and state of the art ResNet50. To further improve the performance we propose our own convolution neural network (CNN) model for Bengali grapheme classification with validation root accuracy 95.32\%, vowel accuracy 98.61\%, and consonant accuracy 98.76\%. We also explore Region Proposal Network (RPN) using VGGNet with a limited setting that can be a potential future direction to improve the performance.
Affective computing is very important in the relationship between man and machine. In this paper, a system for speech emotion recognition (SER) based on speech signal is proposed, which uses new techniques in different stages of processing. The system consists of three stages: feature extraction, feature selection, and finally feature classification. In the first stage, a complex set of long-term statistics features is extracted from both the speech signal and the glottal-waveform signal using a combination of new and diverse features such as prosodic, spectral, and spectro-temporal features. One of the challenges of the SER systems is to distinguish correlated emotions. These features are good discriminators for speech emotions and increase the SER's ability to recognize similar and different emotions. This feature vector with a large number of dimensions naturally has redundancy. In the second stage, using classical feature selection techniques as well as a new quantum-inspired technique to reduce the feature vector dimensionality, the number of feature vector dimensions is reduced. In the third stage, the optimized feature vector is classified by a weighted deep sparse extreme learning machine (ELM) classifier. The classifier performs classification in three steps: sparse random feature learning, orthogonal random projection using the singular value decomposition (SVD) technique, and discriminative classification in the last step using the generalized Tikhonov regularization technique. Also, many existing emotional datasets suffer from the problem of data imbalanced distribution, which in turn increases the classification error and decreases system performance. In this paper, a new weighting method has also been proposed to deal with class imbalance, which is more efficient than existing weighting methods. The proposed method is evaluated on three standard emotional databases.
Adoption of machine learning models in healthcare requires end users' trust in the system. Models that provide additional supportive evidence for their predictions promise to facilitate adoption. We define consistent evidence to be both compatible and sufficient with respect to model predictions. We propose measures of model inconsistency and regularizers that promote more consistent evidence. We demonstrate our ideas in the context of edema severity grading from chest radiographs. We demonstrate empirically that consistent models provide competitive performance while supporting interpretation.
This paper compares well-established Convolutional Neural Networks (CNNs) to recently introduced Vision Transformers for the task of Diabetic Foot Ulcer Classification, in the context of the DFUC 2021 Grand-Challenge, in which this work attained the first position. Comprehensive experiments demonstrate that modern CNNs are still capable of outperforming Transformers in a low-data regime, likely owing to their ability for better exploiting spatial correlations. In addition, we empirically demonstrate that the recent Sharpness-Aware Minimization (SAM) optimization algorithm considerably improves the generalization capability of both kinds of models. Our results demonstrate that for this task, the combination of CNNs and the SAM optimization process results in superior performance than any other of the considered approaches.
The sewerage infrastructure is one of the most important and expensive infrastructures in modern society. In order to efficiently manage the sewerage infrastructure, automated sewer inspection has to be utilized. However, while sewer defect classification has been investigated for decades, little attention has been given to classifying sewer pipe properties such as water level, pipe material, and pipe shape, which are needed to evaluate the level of sewer pipe deterioration. In this work we classify sewer pipe defects and properties concurrently and present a novel decoder-focused multi-task classification architecture Cross-Task Graph Neural Network (CT-GNN), which refines the disjointed per-task predictions using cross-task information. The CT-GNN architecture extends the traditional disjointed task-heads decoder, by utilizing a cross-task graph and unique class node embeddings. The cross-task graph can either be determined a priori based on the conditional probability between the task classes or determined dynamically using self-attention. CT-GNN can be added to any backbone and trained end-to-end at a small increase in the parameter count. We achieve state-of-the-art performance on all four classification tasks in the Sewer-ML dataset, improving defect classification and water level classification by 5.3 and 8.0 percentage points, respectively. We also outperform the single task methods as well as other multi-task classification approaches while introducing 50 times fewer parameters than previous model-focused approaches. The code and models are available at the project page http://vap.aau.dk/ctgnn
Analysing expressions on the person's face plays a very vital role in identifying emotions and behavior of a person. Recognizing these expressions automatically results in a crucial component of natural human-machine interfaces. Therefore research in this field has a wide range of applications in bio-metric authentication, surveillance systems , emotion to emoticons in various social media platforms. Another application includes conducting customer satisfaction surveys. As we know that the large corporations made huge investments to get feedback and do surveys but fail to get equitable responses. Emotion & Gender recognition through facial gestures is a technology that aims to improve product and services performance by monitoring customer behavior to specific products or service staff by their evaluation. In the past few years there have been a wide variety of advances performed in terms of feature extraction mechanisms , detection of face and also expression classification techniques. This paper is the implementation of an Ensemble CNN for building a real-time system that can detect emotion and gender of the person. The experimental results shows accuracy of 68% for Emotion classification into 7 classes (angry, fear , sad , happy , surprise , neutral , disgust) on FER-2013 dataset and 95% for Gender classification (Male or Female) on IMDB dataset. Our work can predict emotion and gender on single face images as well as multiple face images. Also when input is given through webcam our complete pipeline of this real-time system can take less than 0.5 seconds to generate results.
The dependability of power electronics systems, such as three-phase inverters, is critical in a variety of applications. Different types of failures that occur in an inverter circuit might affect system operation and raise the entire cost of the manufacturing process. As a result, detecting and identifying inverter problems for such devices is critical in industry. This study presents the short-time Fourier transform (STFT) for fault classification and identification in three-phase type, voltage source inverter (VSI) switches. Time-frequency representation (TFR) represents the signal analysis of STFT, which includes total harmonic distortion, instantaneous RMS current, RMS fundamental current, total non harmonic distortion, total waveform distortion and average current. The features of the faults are used with a rule-based classifier based on the signal parameters to categorise and detect the switch faults. The suggested method's performance is evaluated using 60 signals containing short and open circuit faults with varying characteristics for each switch in VSI. The classification results demonstrate the proposed technique is good to be implemented for VSI switches faults classification, with an accuracy classification rate of 98.3%.
It is a practical research topic how to deal with multi-device audio inputs by a single acoustic scene classification system with efficient design. In this work, we propose Residual Normalization, a novel feature normalization method that uses frequency-wise normalization % instance normalization with a shortcut path to discard unnecessary device-specific information without losing useful information for classification. Moreover, we introduce an efficient architecture, BC-ResNet-ASC, a modified version of the baseline architecture with a limited receptive field. BC-ResNet-ASC outperforms the baseline architecture even though it contains the small number of parameters. Through three model compression schemes: pruning, quantization, and knowledge distillation, we can reduce model complexity further while mitigating the performance degradation. The proposed system achieves an average test accuracy of 76.3% in TAU Urban Acoustic Scenes 2020 Mobile, development dataset with 315k parameters, and average test accuracy of 75.3% after compression to 61.0KB of non-zero parameters. The proposed method won the 1st place in DCASE 2021 challenge, TASK1A.
The integration of artificial intelligence into clinical workflows requires reliable and robust models. Among the main features of robustness is repeatability. Much attention is given to classification performance without assessing the model repeatability, leading to the development of models that turn out to be unusable in practice. In this work, we evaluate the repeatability of four model types on images from the same patient that were acquired during the same visit. We study the performance of binary, multi-class, ordinal, and regression models on three medical image analysis tasks: cervical cancer screening, breast density estimation, and retinopathy of prematurity classification. Moreover, we assess the impact of sampling Monte Carlo dropout predictions at test time on classification performance and repeatability. Leveraging Monte Carlo predictions significantly increased repeatability for all tasks on the binary, multi-class, and ordinal models leading to an average reduction of the 95% limits of agreement by 17% points.
Graph Neural Networks have become one of the indispensable tools to learn from graph-structured data, and their usefulness has been shown in wide variety of tasks. In recent years, there have been tremendous improvements in architecture design, resulting in better performance on various prediction tasks. In general, these neural architectures combine node feature aggregation and feature transformation using learnable weight matrix in the same layer. This makes it challenging to analyze the importance of node features aggregated from various hops and the expressiveness of the neural network layers. As different graph datasets show varying levels of homophily and heterophily in features and class label distribution, it becomes essential to understand which features are important for the prediction tasks without any prior information. In this work, we decouple the node feature aggregation step and depth of graph neural network, and empirically analyze how different aggregated features play a role in prediction performance. We show that not all features generated via aggregation steps are useful, and often using these less informative features can be detrimental to the performance of the GNN model. Through our experiments, we show that learning certain subsets of these features can lead to better performance on wide variety of datasets. We propose to use softmax as a regularizer and "soft-selector" of features aggregated from neighbors at different hop distances; and L2-Normalization over GNN layers. Combining these techniques, we present a simple and shallow model, Feature Selection Graph Neural Network (FSGNN), and show empirically that the proposed model achieves comparable or even higher accuracy than state-of-the-art GNN models in nine benchmark datasets for the node classification task, with remarkable improvements up to 51.1%.
Histopathological analysis is the present gold standard for precancerous lesion diagnosis. The goal of automated histopathological classification from digital images requires supervised training, which requires a large number of expert annotations that can be expensive and time-consuming to collect. Meanwhile, accurate classification of image patches cropped from whole-slide images is essential for standard sliding window based histopathology slide classification methods. To mitigate these issues, we propose a carefully designed conditional GAN model, namely HistoGAN, for synthesizing realistic histopathology image patches conditioned on class labels. We also investigate a novel synthetic augmentation framework that selectively adds new synthetic image patches generated by our proposed HistoGAN, rather than expanding directly the training set with synthetic images. By selecting synthetic images based on the confidence of their assigned labels and their feature similarity to real labeled images, our framework provides quality assurance to synthetic augmentation. Our models are evaluated on two datasets: a cervical histopathology image dataset with limited annotations, and another dataset of lymph node histopathology images with metastatic cancer. Here, we show that leveraging HistoGAN generated images with selective augmentation results in significant and consistent improvements of classification performance (6.7% and 2.8% higher accuracy, respectively) for cervical histopathology and metastatic cancer datasets.