About this Event
GRADnet Workshop: The Application of Neural networks to Image Recognition
Who: PhD students and postdoctoral researchers interested in modern data science and AI
What: A 1-day workshop that consists of pre-session material (in a github repository)
and tutorials. The participants will have a chance to browse through all the materials
prior to the session and then have the opportunity to get some hands-on experience
during the in-person workshop.
When: Monday, 20 January 2024, 10:00-17:00
Where: University of Portsmouth, FTC Computer Lab
Numbers: 25
In this course, we discuss the most common types of neural networks widely
used in Image Recognition. We will briefly go through the details of how neural networks
learn. We will discuss feeding forward, backpropagation, and gradient descent type
methods.
The focus of the workshop is the implementation (in Python, PyTorch) of these neural networks and applying them to classify images. We will first build some simple networks
with a few convolutional layers, then we proceed by exploiting popular deep neural
networks (e.g., ResNet, Inception https://pytorch.org/vision/stable/models.html).We will discuss the transfer-learning and hyper-parameters tuning.
In our work we use a few Kaggle image dataset, in particular, we aim to recognise
emotions from the images of people's faces, hence we will need the right dataset for
this.
At the end of the session, we will briefly discuss the generative models for image generation, in particular, we focus on the Latent Diffusion model Kadinsky available from Hugging Face (https://huggingface.co/docs/diffusers/en/api/pipelines/kandinsky_v22).
The course will be using PyTorch, however, we will also mention how this
can be done in Keras with TensorFlow backend.
Event Venue & Nearby Stays
University of Portsmouth, Winston Churchill Avenue, Southsea, United Kingdom
GBP 0.00