AI and computer vision

I develop artificial intelligence and computer vision algorithms that turn images into measurements. I also bring AI into 3D and interactive applications: avatars and virtual assistants that listen, talk and respond.

Image and material analysis

At the Italian Institute of Technology I developed algorithms that measure the weathering of excavated archaeological glass and separate biological growth from original pigment in ancient frescoes. Where conservators had a judgement call, they now have a repeatable measurement.

The same approach applies wherever images must be classified and measured: cultural heritage, research, visual inspection of materials.

AI-driven virtual assistants and avatars

• For AGS I led the Unity front end of a life-size, AI-driven deepfake virtual assistant with speech recognition, including a 3D blendshape version.
• At Metide I developed an augmented reality virtual assistant with AI and speech recognition, such as Editem.

Why work with me

I combine the algorithmic side with the visual and interactive one, so I can build both the model and the application people use it through, in Unity, on the web or on dedicated devices. I work with research institutes, where results must be verifiable and documented.