Latest Articles for Deep Learning

New model improves accuracy in segmenting tumors from PET/CT scans.

HESSO simplifies model compression, making neural networks more efficient without losing performance.

A new method for training CNNs directly on mobile and edge devices.

A novel framework enhances ultrasound imaging accuracy during labor monitoring.

New metrics help assess SSL methods using unlabelled data effectively.

Seg-HGNN enhances image segmentation using hyperbolic graph neural networks.

This article reveals methods to interpret CLIP-like models in AI.

MENSA enhances survival predictions by handling multiple events and censoring effectively.

This study proposes seam carving to improve image classification in CNNs.

This framework improves segmentation performance with fewer examples and efficient model usage.

A tool helps users confirm if their data was used in deep learning models.

A novel approach to assess quality in brain MRI image generation.

KRDistill enhances knowledge distillation by addressing data imbalance issues.

A look into how operator learning is shaping modern mathematical modeling.

This study examines the impact of regularization methods on actor networks in offline RL.

Study shows staining variations hinder deep neural networks in cancer prediction.

A new framework improves model learning in 3D with exemplar selection.

A new method improves retinal images for better disease diagnosis.

A new method for identifying AI-generated fake images with improved accuracy.

Exploring Sparse Neural Networks and their performance with challenging training data.

New method improves clustering for large datasets using Stochastic Quantization.

Introducing S-STE, a novel approach to improve sparse neural network training efficiency.

A new method enhances medical image segmentation by reducing information loss.

A new method offers efficient solutions for integral equations using neural networks.

Explore adaptive sampling for better neural network performance with symmetric data.

A new method predicts DNN performance on various hardware quickly and accurately.

A new method improves feature transfer in implicit neural representations for images.

Learn how AMRF enhances image segmentation in industrial applications.

AMBER outperforms CNNs in classifying hyperspectral images, showcasing improved accuracy.

Learn how deep learning is changing our understanding of protein behavior.

New method enhances polyp detection accuracy during colonoscopy procedures.

KAT improves deep learning by using advanced KANs to replace MLPs.

This work enhances CLIP's accuracy by addressing intra-modal overlap using lightweight adapters.

Introducing ComFNO, a method for tackling complex differential equations using deep learning.

WaveMixSR-V2 transforms low-resolution images into high-quality outputs efficiently.

Introducing PAD-FT, a lightweight method to fight backdoor attacks without clean data.

AdEMAMix improves training efficiency by balancing recent and past gradients.

StableMamba enhances image and video processing with improved robustness and performance.

This study evaluates low-latency methods for improving speech quality in noisy conditions.

A study on how technical debt affects deep learning projects.