Hi everyone! I am Juan Borrego-Carazo a predoctoral fellow at Universitat Autònoma de Barcelona, Spain. I am currently working on deep learning for resource constrained devices. Specifically, my current projects are about bronchoscopy tracking and unsupervised semantic segmentation.

My expertise is with deep neural network development, specialized in optimization and deployment: quantization, pruning, NAS, distillation and efficient operations. If you are interested do not doubt to contact me 😄

👔 Linkedin | ✉️ Email | 💬 Issue Me about everything! | 🐦Twitter

• torch2cmsis: Framework for converting neural networks developed with PyTorch to CMSIS-NN.
• SpArSeMod: Framework for microcontroller oriented and automatic neural networks.
• GPyTorch Arc Kernel Implementation: Implementation of Arc Kernel, a kernel for conditional space modelling in Gaussian Processes. Merged PR
• FLOP evaluation of RNNs in PyTorch: Count of FLOP in Recurrent Neural Networks for latency approximation Merged PR

You can check my current pulished articles and conferences at 📗 Research Gate or 📖 Google Scholar

You can also more about what I do in my blog: experiments, learning, development and beyond 📝 BCJuan.

• Unsupervised Semantic Segmentation

• Open to collaborations and new projects