Barcelona Artificial Intelligence Lab

I've trained CNNs, Transformers, GANs, Adversarial Networks, Facenets ... Mostly in Pytorch and sometimes with TF, especially if the solution needs to be optimized to work in the edge (TFlite)

Every projects needs its particular solution, deeplearning is a tool and not an end. Outputs and inputs are always pre and post processed. I like to use data science and data processing tools.

Data is the key to success in deeplearning . Understanding the data creates useful datasets. I have a team of taggers and particular infrastructures to create customs datasets.

Startups need to validate ideas and solutions before going to market. Innovation projects from big companies need to acquire the necessary budget. I can help find how to do it in a reasonable time.

If your team have problems with their projects, bottlenecks or need to optimize their pipelines. As an ex CTO I have years of computer management experience in business. I'm also open to do training.

We build object detection or classification datasets also segmentation (instance and semantic). Sometimes with the help of professionals like doctors. We collect images and use smart ways to maximize the performance of our work.