danieltudosiu Goto Github PK

We are expanding our team

• Two full-time positions - VLMs + Diffusion Models / LLMs / Multimodal / Multi-plane representations
• Three intern positions (both Research Scientist and Research Engineering) - Familiarity with VLM / PEFT / Diffusion Models

• Multi-layer Image and Video Decomposition and Generation
• Controllable and Editable Generativel Modelling
• Interleaved Text and Image Generation

Researched One/Few-Shot(s) Active Learning Segmentation pipelines and how at-scale training can help their generalisation.

Developing state-of-the-art generative models of the human brain and applying them for anomaly detection and disease progression modelling.

• We made the DALL-E of the human brain at MICCAI 2022.
• We created the first morphologically correct generative model of the human brain at MICCAI 2022.
• We extensively analised the usage of VQ-VAE + Transformer pipeline for anomaly detection. Work was published at MIDL and MICCAI.

Working for a greener future by:

• Developing Remote-Sensing Deep Learning models aimed at quantifying current state and future potential of cities' green roof.
• Creating insightful business inteligence for local authorities to decrease their carbon footprint.