Latest Articles for Deep Learning

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.

Pool Skip aids deep networks by addressing elimination singularities during training.

A new method enhances data cleaning for underwater mapping tools.

FDIN enhances video inpainting detection using advanced techniques.

Introducing PureDiffusion to enhance defense mechanisms against backdoor threats.

A multi-domain model improves segmentation of axons and myelin in diverse images.

GCC-UNet enhances retinal vessel segmentation for better disease diagnosis.

This study analyzes the performance of KANs in continual learning using the MNIST dataset.

Two models forecast groundwater levels with improved accuracy through weather data analysis.

A new method for video denoising that improves quality without paired datasets.

A new method enhances liver vessel segmentation in medical imaging.

Researchers improve tumor segmentation in PET/CT images using automated methods.

Studying glottal area segmentation for better medical imaging analysis.

Region Mixup enhances training data diversity for better model performance.

A new model enhances tumor segmentation in PET/CT scans for cancer diagnosis.

A new method enhances metal part monitoring during LPBF using deep learning for image improvement.

This approach improves efficiency and accuracy in analyzing data from multiple devices.

A new method combining models to improve unsupervised domain adaptation in segmentation tasks.

AXE improves model performance while minimizing overflow in accumulator-aware quantization.

A new approach for clearer visualization and understanding of deep learning models.

This study benchmarks DG algorithms to tackle domain shift in medical image analysis.