Search

Modern natural language processing (NLP) state-of-the-art (SoTA) deep learning (DL) models have hundreds of millions of parameters, making them extremely complex. Large datasets are required for training these models, and while pretraining has reduced this requirement, human-labelled datasets are still necessary for fine-tuning. Few-shot learning (FSL) techniques, such as meta-learning, try to train models from smaller datasets to mitigate this cost. However, the tasks used to evaluate these meta-learners frequently diverge from the problems in the real world that they are meant to resolve. This work aims to apply meta-learning to a problem that is more pertinent to the real world: class incremental learning (IL). In this scenario, after completing its training, the model learns to classify newly introduced classes. One unique quality of meta-learners is that they can generalise from a small sample size to classes that have never been seen before, which makes them especially useful for class incremental learning (IL). The method describes how to emulate class IL using proxy new classes. This method allows a meta-learner to complete the task without the need for retraining. To generate predictions, the transformer-based aggregation function in a meta-learner that modifies data from examples across all classes has been proposed. The principal contributions of the model include concurrently considering the entire support and query sets, and prioritising attention to crucial samples, such as the question, to increase the significance of its impact during inference. The outcomes demonstrate that the model surpasses prevailing benchmarks in the industry. Notably, most meta-learners demonstrate significant generalisation in the context of class IL even without specific training for this task. This paper establishes a high-performing baseline for subsequent transformer-based aggregation techniques, thereby emphasising the practical significance of meta-learners in class IL. [ABSTRACT FROM AUTHOR]

Waste management has been an issue due to low awareness among people of any country to lead major environmental contamination, tragic accidents, and unfavorable working conditions for landfill workers. The Lack of precise and efficient object detection could be a barrier in the growth of computer vision-based systems. As per the latest research articles, pre-trained models could be used for Trash Bin detection in real time and for recommending appropriate actions after detection. Using a unique validation dataset made up of predicted trash items, the two classes of acceptable object identification models, YOLO (You Only Look Once) and SSD (Single Shot Multibox Detector), are then contrasted. It is concluded that SSD performs noticeably better than YOLO in identifying trash objects based on several performance metrics computed utilizing multiple open-source research projects. The model is then built up to recognize several trash object types after being pre-trained using Microsoft's COCO (Common Objects in Context) dataset. Our initiative intends to enhance sustainable waste management, make trash sorting incredibly simple, and guard against serious illnesses and accidents at landfill and garbage disposal sites. [ABSTRACT FROM AUTHOR]

The architecture of REEGNet (Recurrent EEG Network) combines convolutional blocks from EEGNet with long short-term memory (LSTM) layers in order to improve temporal modeling capabilities. Additionally, the use of depthwise convolutions enables REEGNet to learn spatial filters specific to distinct frequency information. When evaluated on the exceptionally challenging IMAGENET configuration of the MindBigData Brain Dataset, consisting of 70,060 samples across 5 channels sampled at 128 Hz, REEGNet outperformed state-of-the-art models including EEGNet and LSTM architectures across all class configurations ranging from 2 to 40 classes. Specifically, REEGNet surpasses EEGNet by a margin of 10.6% in multi-class classification scenarios such as configurations with 10 classes, achieving an accuracy of 27.9%. These results highlight REEGNet's potential to increase the accessibility of electroencephalography (EEG) analysis by allowing for reliable classification without extensive computational resources. This makes REEGNet uniquely suited for deployment in resource-constrained real-world environments and applications. [ABSTRACT FROM AUTHOR]

The Current Study elaborates the classification of Lung Cancer by traversing the potential consumption of a Hybrid Model in which ROI (Region of interest) is recognised through different segmentation techniques by extracting key points through different features extraction and optimization techniques are implemented to the dataset of CT scan images of 800 pictures. A total of 800 patients were included in this retrospective study and divided into the training set and test set with a ratio of 8:2 randomly. The research work developed a Hybrid Model by using distinct Optimization techniques i.e. Particle Swarm Optimization (PSO). Artificial BEE colony (ABC) and features extraction method namely SIFT (Scale Invariant Feature Transform, Speed Up Robust Feature (SURF), in context of the least execution time with least mean square error rate with support vector machine (SVM) classifier. Additionally, working of the Hybrid Model has been gauged in respect of parameter Accuracy, Error rate, Precision, Recall, and Execution Time. The overall Accuracy of the hybrid model is 99.56% while recall value is 89%, F-measure 93.44% have been obtained for the Hybrid Model. [ABSTRACT FROM AUTHOR]