This research creates and classifies datasets for Setswana and Sepedi news headlines.
problem Challenges in creating and preparing datasets for low-resourced languages.
method Investigates an approach for data augmentation tailored to low resource languages.
result Improves classification performance on news topic classification task.
New models improve morpheme segmentation in low-resource languages.
problem Improving morpheme segmentation in low-resource languages.
method Two new models: LSTM pointer-generator and sequence-to-sequence with hard monotonic attention.
result Novel models outperform existing ones by up to 11.4% accuracy in low-resource settings.
Improves audio transcription on scarce data with factorized tasks.
problem Weakly labelled data and lack of training samples.
method Factorizing audio transcription into multiple tasks and training a stacked CNN-RNN model.
result Different training methods for intermediate tasks have varying advantages and disadvantages.
LRS-DAG improves domain adaptation for low-resource settings.
problem Maintaining performance on source domain after target domain adaptation.
method Adds encoder layers to map target to source domain, maintaining source performance.
result Outperforms fine-tuning on synthetic low-resource datasets.
This paper explores semi-supervised and transfer learning for low-resource sentiment classification.
problem Training models with small labeled datasets is challenging, especially for new languages and cultures.
method The paper investigates semi-supervised and transfer learning methods to enhance sentiment classification on low-resource datasets.
result The proposed methods significantly improve model performance, particularly with limited training data.
This paper uses GANs to augment sentiment analysis datasets.
problem Lack of data for sentiment analysis.
method Trains GAN models on low-resource datasets for data augmentation.
result GAN-generated data improves sentiment classifier performance.
Meta-learning improves GNN initializations for low-resource drug discovery.
problem Limited labeled data hinders deep learning in drug discovery.
method Model-Agnostic Meta-Learning (MAML) and its variants for graph neural networks initializations.
result Meta-initializations outperform multi-task pre-training baselines on 16 out of 20 tasks and all out-of-distribution tasks.
Study introduces KorFinMTEB for Korean financial texts, revealing model limitations.
problem Limited evaluation benchmarks for low-resource domains, especially Korean.
method Developed KorFinMTEB, a tailored benchmark for Korean financial texts.
result Models perform better on translated benchmarks than on domain-specific ones.
Investigates upsampling vs. upweighting for balanced training on skewed datasets.
problem Balancing training on heavily imbalanced datasets with scarce data.
method Theoretical and empirical analysis of upsampling and upweighting strategies.
result Upsampling and upweighting diverge under stochastic gradient descent, with upsampling leading to faster convergence but higher overfitting risk.
Multilingual LM improves low-resource language modeling.
problem Lack of data for many languages and domains.
method Jointly trained multilingual neural language model with shared parameters.
result Significant improvements in conversational data domain with limited training data.
Detects out-of-domain cases with limited training data.
problem Detecting out-of-domain cases with insufficient in-domain training data.
method Proposes an OOD-resistant Prototypical Network.
result Outperforms state-of-the-art methods in zero-shot OOD detection.
Improves domain adaptation performance in low-resource settings.
problem Training models with scarce data in target domains.
method Augmented cyclic adversarial learning with task-specific cycle-consistency.
result Significant improvement in digit classification and speech recognition tasks.
SCENE-Net improves 3D point cloud segmentation with low resource usage and transparency.
problem Lack of resources and transparency in 3D semantic segmentation models.
method SCENE-Net uses signature shapes identified via GENEOs to achieve semantic segmentation with minimal resources.
result SCENE-Net achieves comparable IoU to state-of-the-art methods with less data and computational resources.
Unified model for sequence labeling and classification.
problem Efficiently perform multiple sequence labeling tasks.
method Generative framework with shared natural language output space.
result Significant improvements in few-shot and low-resource slot labeling.
Paper aims to improve education online in South Africa using NMT for Setswana.
problem Inclusion of machine translation for low-resourced local languages in online education.
method Demonstrates state-of-the-art performance on English-to-Setswana translation using the Transformer architecture.
result Current NMT techniques show promise for African languages, beating previous methods by 5.33 BLEU points.
Improved neural network performance with noisy data.
problem Training neural networks on noisy, automatically annotated data.
method Added a noise layer to a neural network architecture to model and handle noise.
result Improved performance by up to 35% on a low-resource NER task.
This paper tackles rare word problem in low-resource language pairs using NMT.
problem Rare word problem in neural machine translation, especially for low-resource languages.
method Three solutions: enhanced source context, morphology learning, and wordnet synonyms.
result Significant improvements in BLEU scores (+1.0 points) on English-Vietnamese and Japanese-Vietnamese.
FLAML automates model selection and hyperparameter tuning with low resource cost.
problem Automating model selection and hyperparameter tuning for ad-hoc datasets and metrics.
method Conducts trials of different configurations on training data, optimizing for low computational cost.
result Significantly outperforms top-ranked AutoML libraries under smaller budget constraints.
Adaptive scheduling improves multilingual neural machine translation models.
problem Training models on multiple tasks with uniform or proportional sampling leads to poor performance trade-offs.
method Exploring non-adaptive and adaptive task scheduling, including implicit schedules.
result Adaptive schedules improve model performance for low-resource tasks without negatively affecting high-resource tasks.
This paper improves translation of rare words in under-resourced languages.
problem Translation of out-of-vocabulary words in low-resource languages.
method Developed neural and non-neural context models for target-language context integration.
result Neural language models with wide context are most effective in disambiguating OOV word translations.
Method trains deep neural networks on weakly labeled audio data efficiently.
problem Limited training data and lack of temporal labels for audio event detection.
method Multi-instance learning with a new loss function for stacked CNN-RNN.
result Improved performance on low-resource audio datasets.
NukeBERT improves performance on nuclear domain Q&A with less training data.
problem Lack of annotated data for nuclear domain Q&A.
method Developed NQuAD dataset and NukeBERT model incorporating novel BERT vocabulary technique.
result NukeBERT outperformed BERT significantly on NQuAD.
The paper presents a neural machine translation system for Wolof.
problem Low-resource conditions for translating French to Wolof.
method Developed a bidirectional LSTM-based encoder-decoder architecture with attention mechanisms.
result Best model achieved a 47% BLEU score for French-Wolof translation.
Auto-sizing improves machine translation efficiency by 3.9 BLEU points with fewer parameters.
problem Optimizing neural network architecture and hyperparameters for low-resource machine translation is computationally expensive.
method Auto-sizing uses regularization to dynamically adjust network size during training.
result Auto-sizing improves BLEU scores by up to 3.9 points with one-third fewer parameters.
EEG-TCNet improves MI-BMIs with high accuracy and low resource usage.
problem Improving motor-imagery brain-machine interfaces with high accuracy and low resource usage.
method Proposes EEG-TCNet, a novel TCN for embedded MI-BMIs.
result EEG-TCNet achieves 83.84% classification accuracy on MOABB, outperforming SoA by 0.25.
Transformers trained with three normalization changes outperform state-of-the-art on low-resource translation tasks.
problem Improving the training of Transformers, especially on low-resource datasets.
method Three normalization changes: PreNorm, ScaleNorm, and FixNorm.
result Significant improvements in BLEU scores on low-resource translation tasks.
End-to-end Sanskrit TTS developed with limited data, achieving good quality.
problem Developing natural-sounding speech for Sanskrit with scarce data.
method Fine-tuning Tacotron2 model with WaveGlow and transfer learning.
result Achieved an overall MOS of 3.38 from 37 evaluators.
Unified framework explains few-shot multimodal medical imaging performance.
problem Limited labeled data in rare diseases and low-resource settings.
method PAC learning, VC theory, PAC Bayesian analysis, information gain, Chain of Thought reasoning.
result Unified theoretical framework for few-shot multimodal medical imaging.
CNNs improve medical image classification with few samples.
problem Classifying medical images with limited training data.
method Transfer learning using CNNs, representation extraction, and a novel metric for performance prediction.
result CNN-based transfer learning outperforms feature-based methods with high correlation to test set performance.
Deep learning for smile detection on embedded systems.
problem Real-time smile detection in low-resource environments.
method Comparison of neural network architectures on NVidia Jetson platform with asynchronous multithreading.
result Low complexity architectures can achieve similar performance to larger networks with less computation.
New method uses artificial data to improve biomedical text tagging.
problem Lack of labeled data for biomedical tasks.
method Fine-tuning a pre-trained encoder with additional artificial labeled instances.
result Achieved state-of-the-art performance in PICO annotation.
Survey on using large models to train smaller datasets in NLP.
problem Lack of large datasets and computing resources for NLP tasks.
method Analysis of recent transfer learning approaches in NLP.
result Increased demand for transfer learning in NLP due to large models.
Generative Adversarial Networks improve affective speech feature generation.
problem Improving feature representation for emotion recognition.
method Experimented with GAN architectures to generate feature vectors corresponding to emotions.
result GANs generate realistic synthetic samples for emotion recognition.
Study shows multilingual LLM calibration effects improve model confidence but not accuracy.
problem Improving multilingual language model calibration in low-resource settings.
method Analysis of two multilingual benchmarks using instruction-tuning and label smoothing.
result Model confidence increases in low-resource languages after instruction-tuning but accuracy improvements are marginal.
Deep learning predicts hypoxemia from SpO2 data better than anesthesiologists.
problem Improving hypoxemia prediction for better patient outcomes.
method Trained deep learning model on SpO2 data to predict hypoxemia.
result Deep learning model outperforms anesthesiologists in predicting hypoxemia.
Generative models predict page quality without training, useful for low-resource settings.
problem Detecting low-quality content in web articles.
method Human evaluation and analysis of 500 million web articles.
result Generative models can predict page quality without training, useful for low-resource settings.
Paper proposes a new speech representation benchmark and model.
problem Lack of benchmarks for comparing speech representations.
method Unsupervised triplet-loss objective for training a universal non-semantic speech representation.
result Proposed representation outperforms other models on benchmark and transfer learning tasks.
Model improves classification with scarce data by focusing on relevant parts of inputs.
problem Classification tasks with limited training data.
method Approximates exact match with probability computation, learns relevant parts, boosts errors.
result Remarkable results on text classification, benefits in balanced and unbalanced cases.
Trains neural machine translation models for 5 Southern African languages.
problem Lack of shared resources and reproducible results for African languages.
method Trained neural machine translation models on publicly-available datasets.
result Provides code for training and evaluating models, aiming to spur future research.
Paper proposes a new method for better POS tagging adaptation.
problem Difficulty of pre-trained units learning target-specific patterns.
method Augment target-network with normalised, weighted, and randomly initialized units.
result Achieves state-of-the-art performances on POS tagging.
CNN-DTW system improves keyword spotting in under-resourced languages.
problem Keyword spotting in nearly zero-resource languages.
method Multilingual bottleneck features, CNN-DTW, DTW template matching, convolutional neural network.
result Multilingual BNFs improve CNN-DTW by 10.9%.
Improved neural NER by optimizing large corpora for German.
problem Low-resource language named entity recognition.
method Optimized large corpora, lemmatization, part-of-speech tagging, and detailed optimization.
result Up to 11% improvement in F-score on German NER tasks.
Neural machine translation models trained for 5 South African languages.
problem Lack of resources and research for machine translation in African languages.
method Training neural machine translation models for 5 South African languages using modern techniques.
result Promises of neural machine translation for African languages.
OT domain adaptation improves aphasia detection across languages.
problem Detecting aphasia in low-resource languages with limited data.
method Utilized OT domain adaptation to map linguistic features across multiple languages.
result OT domain adaptation significantly improved F1 scores for French and Mandarin aphasia detection.
Paper improves natural language understanding with less data using a new training method.
problem Limited data hinders performance of small models in natural language tasks.
method Generation-Distillation: uses large finetuned models to generate new training data and distill knowledge into smaller models.
result Achieves comparable performance to BERT with 300x fewer parameters and outperforms prior distillation methods.
FMT model improves multimodal sequential learning across language, vision, and acoustic data.
problem Modeling spatio-temporal dynamics across multiple modalities.
method Factorized Multimodal Transformer (FMT) that models intramodal and intermodal dynamics in a factorized manner.
result FMT outperforms existing models on 3 datasets and 21 labels, setting new state of the art.
NMIXX fine-tunes embeddings for finance, outperforming general models in Korean.
problem Financial embeddings struggle in low-resource languages like Korean.
method Fine-tuned with 18.8K triplets, hard negatives, and translations.
result NMIXX achieves gains of +0.10 on English FinSTS and +0.22 on KorFinSTS.
Lin-DBSCAN is a fast density-based clustering algorithm for spatial data.
problem Efficient clustering of large datasets without prior knowledge of cluster number and shape.
method Grid-based scan and merge approach to a discrete density model of DBSCAN.
result Lin-DBSCAN outperforms DBSCAN in efficiency and validity for spatial data clustering.