Enhancing computer vision with computational neurosciences
Exploiting the behavior of biological neurons to build explainable solutions.
Using the power of spike-based computation for low energy consumption.
Improving generalization with flexible learning and adaptation.
At Kaptios, we believe that applying computational neurosciences has the power to address modern issues inherent to the use of Deep Learning in decision making systems.
The project
01
Spike-based enhanced medical image analysis
The SynapSentry project firstly aims to tackle the issues exposed by Deep Learning based approaches in medical image analysis. By investigating the use of biologically-plausible neural networks for breast anomaly detection, we aim to provide solutions that meet healthcare providers modern requirements.
02
Towards a global computational neuroscience engineering
Creating the tools that will allow easier and flawless integration of Spiking Neural Networks (SNNs) in APIs is a crucial step in the development of scalable solutions.
03
Losing energy efficiency is no longer an option
While Deep Learning has proven to be powerful in computer vision, its performance comes with a high energy cost and carbon footprint that needs to be addressed. Taking advantage of sparse activation patterns and event-driven computation in SNNs could possibly overcome these issues.
04
AI as a fast, scalable, affordable and green solution
This is the aim of SynapSentry. Providing radiology services with a detection system that has low maintenance costs and can be quickly and easily adapted to new analysis demands.
Our partners
Our company relies on different partnerships that provide us with the ressources needed at each step of the SynapSentry project.
