.

I learned how deep learning works through hands-on exercises in computer vision and natural language processing. I trained deep learning models from scratch, mastering tools and techniques to achieve highly accurate results in tasks such as image recognition and language translation. Additionally, I gained experience leveraging state-of-the-art pre-trained models to save time and accelerate deep learning applications. The workshop taught me how multi-layered artificial neural networks can recognize complex patterns in data that are too subtle for traditional expert-written software.

I explored the principles behind generative models. By the end of the course, I was able to build and train a U-Net to generate images from pure noise, enhance image quality using the denoising diffusion process, and incorporate context embeddings to control image outputs. Additionally, I could generate images from English text prompts by leveraging the Contrastive Language-Image Pretraining (CLIP) neural network, bridging the gap between visual and linguistic representations.