Latest AI and machine learning research in smoking & tobacco for healthcare professionals.
Modern diffusion models encounter a fundamental trade-off between training efficiency and generati...
Recent advances in deep learning-based point cloud registration have improved generalization, yet ...
This protocol outlines a scoping review designed to systematically map the existing body of eviden...
Recent advances in diffusion-based text-to-image generation have demonstrated promising results th...
With transformer-based models and the pretrain-finetune paradigm becoming mainstream, the high sto...
The deployment of computer-aided diagnosis systems for cervical cancer screening using whole slide...
Computational pathology and whole-slide image (WSI) analysis are pivotal in cancer diagnosis and p...
State Space Models (SSMs) have recently emerged as an alternative to Vision Transformers (ViTs) du...
Most of existing blind omnidirectional image quality assessment (BOIQA) models rely on viewport ge...
Dichotomous Image Segmentation (DIS) is a high-precision object segmentation task for high-resolut...
Deep learning has demonstrated great promise in cancer classification from whole-slide images (WSI...
Digital versions of real-world text documents often suffer from issues like environmental corrosio...
Advancing AI in computational pathology requires large, high-quality, and diverse datasets, yet ex...
Accurate segmentation of 3D vascular structures is essential for various medical imaging applicati...
Addictive behaviour is shaped by the dynamic interaction of implicit, bottom-up and explicit, top-do...
Autoregressive (AR) modeling, known for its next-token prediction paradigm, underpins state-of-the...
As technology has become more embedded into our society, the security of modern-day systems is par...
Visual transformers have achieved remarkable performance in image classification tasks, but this p...
Most recently, learned image compression methods have outpaced traditional hand-crafted standard c...
Computational pathology, integrating computational methods and digital imaging, has shown to be ef...
Adversarial attacks pose a significant threat to deep learning models, particularly in safety-crit...