Evaluating rule-based and generative data augmentation techniques for legal document classification

Automated segmentation of dental restorations using deep learning: exploring data augmentation techniques.

The growing success and popularity of deep learning in computer vision have resulted in the availability of several pretrained deep learning architectures (such as AlexNet, ResNet, VGGNet among others). A common practice in deep learning research is to use one of the pre-trained models and fine-tune it to a given target task, using training data from the target. However, training a deep learning model efficiently necessitates expensive, high-quality GPUs and distributed computing infrastructures. Some applications (such as those running on mobile platforms) are severely limited in terms of memory and computational resources; in these applications, it is a significant challenge to fine-tune a pre-trained deep learning model to a target task, using large amounts of target training data. Cloud services can be leveraged for training, but involve issues with data privacy and cost. In such applications, it is important to select an informative subset of the training data and fine-tune the deep model using only the selected subset. In this paper, we propose a novel framework to address this problem. We pose subset selection as a constrained NP-hard integer quadratic programming problem and derive an efficient linear relaxation to select a subset of exemplar instances. Our extensive empirical studies on three challenging vision datasets (from different application domains) using three commonly used pre-trained deep learning models corroborate the potential of our framework for real-world, resource-constrained applications.

In various machine learning tasks, including image classification, data augmentation is widely used to expand the training dataset and prevent overfitting. In traditional data augmentation techniques, images from training data are flipped, distorted, added with some noise, or cropped to create new samples from the original images. This paper presents an improved and more effective data augmentation technique for image classification tasks, called Flip-and-Hide (FnH). Our technique was validated on four well-known datasets, MNIST, FMNIST, CIFAR-10 and CIFAR-100. We found 0.03% and 0.12% improvements on FMNIST and CIFAR-10, respectively, compared to the conventional Hide-and-Seek (HaS) augmentation technique. Moreover, we demonstrated the superiority of the proposed FnH augmentation technique by performing some visualisations. Consequently, our technique is particularly useful for image classification tasks that require limited training data.

Sentiment analysis is a subfield of Natural Language Processing (NLP) which tries to process a text to extract opinions or attitudes towards topics or entities. Recently, the use of deep learning methods for sentiment analysis has received noticeable attention from researchers. Generally, different deep learning methods have shown superb performance in sentiment analysis problem. However, deep learning models are different in nature and have different strengths and limitations. For example, convolutional neural networks are useful for extracting local structures from data, while recurrent models are able to learn order dependence in sequential data. In order to combine the advantages of different deep models, in this paper we have proposed a novel approach for aspect-based sentiment analysis which utilizes deep ensemble learning. In the proposed method, we first build four deep learning models, namely CNN, LSTM, BiLSTM and GRU. Then the outputs of these models are combined using stacking ensemble approach where we have used logistic regression as meta-learner. The results of applying the proposed method on the real datasets show that our method has increased the accuracy of aspect-based prediction by 5% to 20% compared to the basic deep learning methods.

We investigate the impact of dataset augmentation and synthetic generation techniques on the accuracy of supervised audio classification based on state-of-the-art neural networks used as classifiers. Dataset augmentation techniques are applied upon the raw sound and its transformed image format. Specifically, sound augmentation techniques are applied prior to spectral-based transformation and include time stretching, pitch shifting, noise addition, volume controlling, and time shifting. Image augmentation techniques are applied after the transformation of the sound into a scalogram, involving scaling, shearing, rotation, and translation. Synthetic sound generation is based on the AudioGen generative model, triggered through a series of customized prompts. Augmentation and synthetic generation are applied to three sound categories: (a) human sounds, (b) animal sounds, and (c) sounds of things, with each category containing ten sound classes with 20 samples retrieved from the ESC-50 dataset. Sound- and image-orientated neural network classifiers have been used to classify the augmented datasets and their synthetic additions. VGGish and YAMNet (sound classifiers) employ spectrograms, while ResNet50 and DarkNet53 (image classifiers) employ scalograms. The streamlined AI-based process of augmentation and synthetic generation, enhanced classifier fine-tuning and inference allowed for a consistent, multicriteria-comparison of the impact. Classification accuracy has increased for all augmentation and synthetic generation scenarios; however, the increase has not been uniform among the techniques, the sound types, and the percentage of the training set population increase. The average increase in classification accuracy ranged from 2.05% for ResNet50 to 9.05% for VGGish. Our findings reinforce the benefit of audio augmentation and synthetic generation, providing guidelines to avoid accuracy degradation due to overuse and distortion of key audio features.

Maxsmi: Maximizing molecular property prediction performance with confidence estimation using SMILES augmentation and deep learning

Multi-site cross-organ calibrated deep learning (MuSClD): Automated diagnosis of non-melanoma skin cancer☆

Classification of brain tumors is one of the daunting tasks in medical imaging and incorrect decisions during the diagnosis process may lead to increased human fatality. Latest advances in artificial intelligence and deep learning have opened the path for the success of numerous medical image analysis tasks, including the recognition of brain tumors. In this paper, we propose a simple architecture based on deep learning that results in strong generalization without needing much preprocessing. The proposed Multi-modal Squeeze and Excitation model (MSENet) receives multiple representations of a given tumor image, learns end-to-end and effectively predicts the severity level of tumor. Convolution feature descriptors from multiple deep pre-trained models are used to effectively describe the tumor images and are supplied as input to the proposed MSENet. The squeeze and excitation blocks of the MSENet allow the model to prioritize tumor regions while giving less emphasis to the rest of the image, serving as an attention mechanism in the model. The proposed model is evaluated on the benchmark brain tumor dataset that is publicly accessible from the Figshare repository. Experimental studies reveal that in terms of model parameters, the proposed approach is simple and leads to competitive performance compared to those of existing complex models. By increasing complexity, the proposed model leads to generalize better and achieves state-of-the-art accuracy of 96.05% on Figshare dataset. Compared to the existing models, the proposed model neither uses segmentation nor uses augmentation techniques and without much pre-processing achieves competitive performance.

Deep language models, like ELMo, BERT and GPT, have achieved impressive results on several natural language tasks. These models are pretrained on large corpora of unlabeled general domain text and later supervisedly trained on downstream tasks. An optional step consists of finetuning the language model on a large intradomain corpus of unlabeled text, before training it on the final task. This aspect is not well explored in the current literature. In this work, we investigate the impact of this step on named entity recognition (NER) for Portuguese legal documents. We explore different scenarios considering two deep language architectures (ELMo and BERT), four unlabeled corpora and three legal NER tasks for the Portuguese language. Experimental findings show a significant improvement on performance due to language model finetuning on intradomain text. We also evaluate the finetuned models on two general-domain NER tasks, in order to understand whether the aforementioned improvements were really due to domain similarity or simply due to more training data. The achieved results also indicate that finetuning on a legal domain corpus hurts performance on the general-domain NER tasks. Additionally, our BERT model, finetuned on a legal corpus, significantly improves on the state-of-the-art performance on the LeNER-Br corpus, a Portuguese language NER corpus for the legal domain.

We present several deep learning models for assessing the morphometric fidelity of deep grey matter region models extracted from brain MRI. We test three different convolutional neural net architectures (VGGNet, ResNet and Inception) over 2D maps of geometric features. Further, we present a novel geometry feature augmentation technique based on parametric spherical mapping. Finally, we present an approach for model decision visualization, allowing human raters to see the areas of subcortical shapes most likely to be deemed of failing quality by the machine. Our training data is comprised of 5200 subjects from the ENIGMA Schizophrenia MRI cohorts, and our test dataset contains 1500 subjects from the ENIGMA Major Depressive Disorder cohorts. Our final models reduce human rater time by 46–70%. ResNet outperforms VGGNet and Inception for all of our predictive tasks.

Platelets play a crucial role in diagnosing and detecting various diseases, influencing the progression of conditions and guiding treatment options. Accurate identification and classification of platelets are essential for these purposes. The present study aims to create a synthetic database of platelet images using Generative Adversarial Networks (GANs) and validate its effectiveness by comparing it with datasets of increasing sizes generated through traditional augmentation techniques. Starting from an initial dataset of 71 platelet images, the dataset was expanded to 141 images (Level 1) using random oversampling and basic transformations and further to 1463 images (Level 2) through extensive augmentation (rotation, shear, zoom). Additionally, a synthetic dataset of 300 images was generated using a Wasserstein GAN with Gradient Penalty (WGAN-GP). Eight pre-trained deep learning models (DenseNet121, DenseNet169, DenseNet201, VGG16, VGG19, InceptionV3, InceptionResNetV2, and AlexNet) and two custom CNNs were evaluated across these datasets. Performance was measured using accuracy, precision, recall, and F1-score. On the extensively augmented dataset (Level 2), InceptionV3 and InceptionResNetV2 reached 99% accuracy and 99% precision/recall/F1-score, while DenseNet201 closely followed, with 98% accuracy, precision, recall and F1-score. GAN-augmented data further improved DenseNet’s performance, demonstrating the potential of GAN-generated images in enhancing platelet classification, especially where data are limited. These findings highlight the benefits of combining traditional and GAN-based augmentation techniques to improve classification performance in medical imaging tasks.

PurposeWe developed new deep learning-based hierarchical brain segmentation (DLHBS) method that can segment T1-weighted MR images (T1WI) into 107 brain subregions and calculate the volume of each subregion. This study aimed to evaluate the repeatability and reproducibility of volume estimation using DLHBS and compare them with those of representative brain segmentation tools such as statistical parametric mapping (SPM) and FreeSurfer (FS).MethodsHierarchical segmentation using multiple deep learning models was employed to segment brain subregions within a clinically feasible processing time. The T1WI and brain mask pairs in 486 subjects were used as training data for training of the deep learning segmentation models. Training data were generated using a multi-atlas registration-based method. The high quality of training data was confirmed through visual evaluation and manual correction by neuroradiologists. The brain 3D-T1WI scan–rescan data of the 11 healthy subjects were obtained using three MRI scanners for evaluating the repeatability and reproducibility. The volumes of the eight ROIs—including gray matter, white matter, cerebrospinal fluid, hippocampus, orbital gyrus, cerebellum posterior lobe, putamen, and thalamus—obtained using DLHBS, SPM 12 with default settings, and FS with the “recon-all” pipeline. These volumes were then used for evaluation of repeatability and reproducibility.ResultsIn the volume measurements, the bilateral thalamus showed higher repeatability with DLHBS compared with SPM. Furthermore, DLHBS demonstrated higher repeatability than FS in across all eight ROIs. Additionally, higher reproducibility was observed with DLHBS in both hemispheres of six ROIs when compared with SPM and in five ROIs compared with FS. The lower repeatability and reproducibility in DLHBS were not observed in any comparisons.ConclusionOur results showed that the best performance in both repeatability and reproducibility was found in DLHBS compared with SPM and FS.

Deep learning is an important technique for extracting value from big data. However, the effectiveness of deep learning requires large volumes of high quality training data. In many cases, the size of training data is not large enough for effectively training a deep learning classifier. Data augmentation is a widely adopted approach for increasing the amount of training data. But the quality of the augmented data may be questionable. Therefore, a systematic evaluation of training data is critical. Furthermore, if the training data is noisy, it is necessary to separate out the noise data automatically. In this paper, we propose a deep learning classifier for automatically separating good training data from noisy data. To effectively train the deep learning classifier, the original training data need to be transformed to suit the input format of the classifier. Moreover, we investigate different data augmentation approaches to generate sufficient volume of training data from limited size original training data. We evaluated the quality of the training data through cross validation of the classification accuracy with different classification algorithms. We also check the pattern of each data item and compare the distributions of datasets. We demonstrate the effectiveness of the proposed approach through an experimental investigation of automated classification of massive biomedical images. Our approach is generic and is easily adaptable to other big data domains.

The impact analysis of deep learning models for COVID-19-infected X-ray images is an extremely challenging task. Every model has unique capabilities that can provide suitable solutions for some given problem. The prescribed work analyzes various deep learning models that are used for capturing the chest X-ray images. Their performance-defining factors, such as accuracy, f1-score, training and the validation loss, are tested with the support of the training dataset. These deep learning models are multi-layered architectures. These parameters fluctuate based on the behavior of these layers, learning rate, training efficiency, or over-fitting of models. This may in turn introduce sudden changes in the values of training accuracy, testing accuracy, loss or validation loss, f1-score, etc. Some models produce linear responses with respect to the training and testing data, such as Xception, but most of the models provide a variation of these parameters either in the accuracy or the loss functions. The prescribed work performs detailed experimental analysis of deep learning image neural network models and compares them with the above said parameters with detailed analysis of these parameters with their responses regarding accuracy and loss functions. This work also analyses the suitability of these model based on the various parameters, such as the accuracy and loss functions to various applications. This prescribed work also lists out various challenges on the implementation and experimentation of these models. Solutions are provided for enhancing the performance of these deep learning models. The deep learning models that are used in the prescribed work are Resnet, VGG16, Resnet with VGG, Inception V3, Xception with transfer learning, and CNN. The model is trained with more than 1500 images of the chest-X-ray data and tested with around 132 samples of the X-ray image dataset. The prescribed work analyzes the accuracy, f1-score, recall, and precision of these models and analyzes these parameters. It also measures parameters such as training accuracy, testing accuracy, loss, and validation loss. Each epoch of every model is recorded to measure the changes in these parameters during the experimental analysis. The prescribed work provides insight for future research through various challenges and research findings with future directions.

The rapid development and progress in deep machine-learning techniques have become a key factor in solving the future challenges of humanity. Vision-based target detection and object classification have been improved due to the development of deep learning algorithms. Data fusion in autonomous driving is a fact and a prerequisite task of data preprocessing from multi-sensors that provide a precise, well-engineered, and complete detection of objects, scene or events. The target of the current study is to develop an in-vehicle information system to prevent or at least mitigate traffic issues related to parking detection and traffic congestion detection. In this study we examined to solve these problems described by (1) extracting region-of-interest in the images (2) vehicle detection based on instance segmentation, and (3) building deep learning model based on the key features obtained from input parking images. We build a deep machine learning algorithm that enables collecting real video-camera feeds from vision sensors and predicting free parking spaces. Image augmentation techniques were performed using edge detection, cropping, refined by rotating, thresholding, resizing, or color augment to predict the region of bounding boxes. A deep convolutional neural network F-MTCNN model is proposed that simultaneously capable for compiling, training, validating and testing on parking video frames through video-camera. The results of proposed model employing on publicly available PK-Lot parking dataset and the optimized model achieved a relatively higher accuracy 97.6% than previous reported methodologies. Moreover, this article presents mathematical and simulation results using state-of-the-art deep learning technologies for smart parking space detection. The results are verified using Python, TensorFlow, OpenCV computer simulation frameworks.