arXiv research

Study examines how classifier performance is affected by training data quality.

Theoretical analysis of data quality and synergies in LLMs.

Federated learning leaks participant dataset quality even with secure aggregation.

The quality of an induced model by a learning algorithm is dependent on the quality of the training data and the hyper-parameters supplied to the learning algorithm. Prior work has shown that improving the quality of the training data (i.e., by removing low quality instances) or tuning the learning algorithm hyper-para…

New measure assesses time series pre-training data quality without labels.

Generative networks improve fluid simulation quality by focusing on high frequencies.

We describe a novel method for training high-quality image denoising models based on unorganized collections of corrupted images. The training does not need access to clean reference images, or explicit pairs of corrupted images, and can thus be applied in situations where such data is unacceptably expensive or impossi…

When the available data of a target speaker is insufficient to train a high quality speaker-dependent neural text-to-speech (TTS) system, we can combine data from multiple speakers and train a multi-speaker TTS model instead. Many studies have shown that neural multi-speaker TTS model trained with a small amount data f…

Improves sample quality of generative models using energy-based methods.

New methods predict neural network quality without access to training data.

Improves GAN sample quality by refining the training dataset.

Private PGB boosts synthetic data quality using GANs and privacy techniques.

Paper proposes a new method to measure model sensitivity using final model only.

New method reduces memorization in diffusion models without sacrificing image quality.

Framework evaluates quality of synthetic data generated with differential privacy.

In many situations, we need to build and deploy separate models in related environments with different data qualities. For example, an environment with strong observation equipments (e.g., intensive care units) often provides high-quality multi-modal data, which are acquired from multiple sensory devices and have rich-…

Federated learning is a distributed form of machine learning where both the training data and model training are decentralized. In this paper, we use federated learning in a commercial, global-scale setting to train, evaluate and deploy a model to improve virtual keyboard search suggestion quality without direct access…

Thanks to the growing availability of spoofing databases and rapid advances in using them, systems for detecting voice spoofing attacks are becoming more and more capable, and error rates close to zero are being reached for the ASVspoof2015 database. However, speech synthesis and voice conversion paradigms that are not…

DsDm selects data to improve model performance, avoiding handpicked notions of quality.

DeepSubQE estimates translation quality for subtitles, improving on existing methods.

PaDGAN generates diverse, high-quality designs with improved performance.

Paper proposes data quality measures for large-scale high-dimensional data.

Pre-training text representations have led to significant improvements in many areas of natural language processing. The quality of these models benefits greatly from the size of the pretraining corpora as long as its quality is preserved. In this paper, we describe an automatic pipeline to extract massive high-quality…

Filtering data with a pre-trained model improves multimodal contrastive learning performance.

Survey examines data quality challenges in edge ML.

Generative adversarial networks (GANs) are highly effective unsupervised learning frameworks that can generate very sharp data, even for data such as images with complex, highly multimodal distributions. However GANs are known to be very hard to train, suffering from problems such as mode collapse and disturbing visual…

Meta-Router optimizes LLM selection using gold-standard and preference-based data.

QA-Token improves tokenization for noisy data, boosting model performance.

Enhances diffusion models by preprocessing data to improve reconstruction quality.

The paper proposes a method to assess and improve data quality using GBDT training dynamics.

Regurgitative training with synthetic data harms LLM performance.

In this paper we propose WaveGlow: a flow-based network capable of generating high quality speech from mel-spectrograms. WaveGlow combines insights from Glow and WaveNet in order to provide fast, efficient and high-quality audio synthesis, without the need for auto-regression. WaveGlow is implemented using only a singl…

Improved StyleGAN model enhances image quality and invertibility.

A method merges two pretrained diffusion experts to improve image quality and likelihood.

One of the biggest bottlenecks in a machine learning workflow is waiting for models to train. Depending on the available computing resources, it can take days to weeks to train a neural network on a large dataset with many classes such as ImageNet. For researchers experimenting with new algorithmic approaches, this is …

Diffusion models generate private synthetic images with high quality.

Training deep neural networks requires massive amounts of training data, but for many tasks only limited labeled data is available. This makes weak supervision attractive, using weak or noisy signals like the output of heuristic methods or user click-through data for training. In a semi-supervised setting, we can use a…

Flow Matching for count data improves sample quality and efficiency.

Automatically assesses the quality of online health articles.

End-to-end Sanskrit TTS developed with limited data, achieving good quality.

VADD enhances discrete diffusion models by capturing inter-dimensional correlations, improving sample quality.

Generative models assess quality on time-series data using ITS and FITD.

A key aspect of automating predictive machine learning entails the capability of properly triggering the update of the trained model. To this aim, suitable automatic solutions to self-assess the prediction quality and the data distribution drift between the original training set and the new data have to be devised. In …

TempBalance boosts model performance with low data.

The time complexity of support vector machines (SVMs) prohibits training on huge data sets with millions of data points. Recently, multilevel approaches to train SVMs have been developed to allow for time-efficient training on huge data sets. While regular SVMs perform the entire training in one -- time consuming -- op…

Improved GEC models use scored data from large pretraining to outperform.

Meta-CoTGAN improves adversarial text generation by preventing mode collapse.

Generative models predict page quality without training, useful for low-resource settings.