SummarizedML

Novel membership inference attacks that leverage the Gaussian neighborhood of a DNN's training data to improve performance. 📄 arxiv.org/abs/2307.05193…

$\beta$D-Bayes, a generalised posterior from a Bayesian posterior from which the data generative process can be sampled 📄 arxiv.org/abs/2307.05194…

Three new reject optionmodels for out-of-distribution setups and define an optimal strategy for OOD setups. 📄 arxiv.org/abs/2307.05199…

A new weakly supervised learning framework for knowledge distillation in video classification that is designed to improve the efficiency and accuracy of the 📄 arxiv.org/abs/2307.05201…

We rethink cross- and self-attention from the viewpoint of traditional feature matching and filtering. 📄 arxiv.org/abs/2307.05180…

An end-to-end Transformer with Co-Attention gaTedVision-Language for VQLA in surgical scenarios. 📄 arxiv.org/abs/2307.05182…

We introduce a novel decorrelation method using Convex Neural OptimalTransport Solvers (Cnots), that is able to decorrel 📄 arxiv.org/abs/2307.05187…

A simple linear regression based approach for learning the weights and biases of a neural network, as an alternative to standard gradient-based back 📄 arxiv.org/abs/2307.05189…

We explore the music adaption of petertoddSSL with two distinctive speech-related models, data2vec1.0 and Hubert, and 📄 arxiv.org/abs/2307.05161…

Parameter Efficient Fine Tuning for Clinical Dialogue Summarization . 📄 arxiv.org/abs/2307.05162…

A mapping study of the state-of-the-art in machine learning methods for estimating the SoH and RUL of lead-acid 📄 arxiv.org/abs/2307.05163…

A novel approach for learning human values that can be represented as a set of arguments. 📄 arxiv.org/abs/2307.05174…

A deep neurological motor primitive model capable of all operations of linear movement primitives. 📄 arxiv.org/abs/2307.05141…

A generative model that can generate multiple samples of any synthetic identity in state-of-the-art pretrained GAN latent spaces 📄 arxiv.org/abs/2307.05151…

Two machine-learning algorithms for selecting events where neutrallong-lived particles decay within the detector volume studying their accuracy and inference time when 📄 arxiv.org/abs/2307.05152…

A modular multimodal architecture that can exploit the attention priors that humans typically have when observing others. 📄 arxiv.org/abs/2307.05158…

This is an overview of the eleventh edition of the BioASQ challenge in the fixmecontext of the Conference and Labs of the Evaluation Forum ( 📄 arxiv.org/abs/2307.05131…

SSL can be used in spontaneous TTS and MOS prediction to further quantify and understand how SSL can beused in spontaneous speech. 📄 arxiv.org/abs/2307.05132…

A new metric based on prompt templates to study the alignment between the content specified in the prompt and the corresponding generatedFINESTimages. 📄 arxiv.org/abs/2307.05134…

Unsupervised sensing algorithms and 3D Augmented Reality visualization for real-time, automatic solar panel detection andrepair . 📄 arxiv.org/abs/2307.05136…