로봇및기계전자공학과

Artificial intelligence (AI) technologies, including machine/deep learning, are greatly advancing robot vision, biomedical engineering, and IT fields.

We are conducting research on AI for image and signal analysis and computer vision for surgical robot navigation.

We are developing core algorithms for image generation, classification, segmentation, tracking, and registration.

In particular, we have recently developed meta-learning methods to enable models to learn with limited data, generative models to synthesize more realistic training data, weakly supervised approaches to reduce the cost of dataset construction, and bias mitigation techniques to prevent the degradation of AI performance towards bias that exists in the data.

Medical imaging technologies have significant potentials in medical diagnosis, early medical intervention, therapy managing plan, and vision system of robot surgery by providing comprehensive physiological information, non-invasively or minimum-invasively.

With the multidisciplinary technologies including artificial intelligence, physics, signal and imaging processing, DGIST RE are developing the next generation medical imaging technologies and systems such as Vision for robotic surgery, Brain-Computer Interface, X-ray & Computed Tomography (CT), Holographic imaging, Optical imaging, and Ultrasound imaging.

Our main goal is to enable a next generation perception systems that can be implemented in robots as well as image processing algorithms for rapid and accurate diseases and brain abnormalities.