Liquid lenses for fast focusing in machine vision

The iToBoS project is usually categorized within the field of medical imaging. However, all the technical solutions used or to be developed in the project, from the optics to the artificial intelligence algorithms, are part of the wider field of machine vision.

Invertible Denoising Network: A Light Solution for Real Noise Removal

The task of real-world noise removal is challenging, as the noise model is highly complex. CNNs have lately achieved state-of-the-art denoising performance, but the networks are extremely large and complicated for a better accuracy.

Diffusion Models

Diffusion Models are a type of likelihood-based models that have recently generated synthetic images of excellent quality.

New scenarios of the teledermatology

In this article we analyze new scenarios and business cases of teledermatology.

Non-contact dermoscopy for the early detection of skin cancer

iToBoS project was presented through a poster titled “Non-contact dermoscopy for the early detection of skin cancer” in the interdisciplinary workshop Cluster of Excellence PhoenixD.

Strengths, Weaknesses and Threats of teledermatology

This article presents an analysis of the strengths, weaknesses and threats of teledermatology.

Primary and Secondary teledermatology

Primary teledermatology means a mobile dermatology linking directly patient and dermatologist, while secondary teledermatology is a way to enlarge the access to specialist care.

Teledermatology: an overview

The main added value of teledermatology is to ensure a wider access to specialist skin care (“Telemedicine is a way of moving patient’s information rather than patients[1]).

Ugly Duckling and Melanoma

One of the key objectives of the iToBoS project is to educate the general public about the most dangerous skin cancer melanoma and the visual signs that could help in early detection.

Introduction to YOLO Object Detection Algorithm

YOLO (You Only Look Once) is an algorithm that uses neural networks to offer real-time detection. Its popularity is due to being much faster than other methods while still providing a good accuracy. It is fast by design: it has fewer convolutional layers (9 instead 24) and fewer filters.