Computer Vision Paris Meetup #28

### The Computer Vision Meetup Paris is back 🔥

For this 28th edition, the Meetup will be held on Tuesday, January, 23, 7p.m at 48 Boulevard des Batignolles 75017 Paris.

As usual, we have 2 special guests, Willy Au & Christian Alaka of Allisone and Vincent Blot & Alexandra Lorenzo of Quantmetry.

🎤 Here are the talks of the evening (the talk are in French 🇫🇷 with English supports 🇬🇧) :

• Pre-training with Diffusion models for Dental Radiography segmentation presented by Willy Au & Christian Alaka

Deep generative models have emerged as one of the most powerful tools in modern AI. While there have been substantial advances and spotlight given to text and image generation, few work focuses on leveraging them for improving downstream computer vision tasks. Our method, “Pre-training with Diffusion models for Dental Radiography segmentation” (Deep Generative Models workshop @ MICCAI 2023, Vancouver) introduces a method to use generative modeling as a pre-training task for label efficient dental x-ray segmentation. This application proves useful in the medical field as pixel-precise xray segmentation, specifically of dental radiography, is highly limited by the cost of labeling which requires specific expertise and labor-intensive annotations. In particular, we will give a quick overview of the research in this field stemming from Denoising Diffusion Probabilistic Models (DDPM), how it is used for medical image segmentation and how we leveraged it in a simple way for training an efficient model with very few annotated images (<10 images). Our proposed method is competitive with other state-of-the-art pre-training methods (MoCo, SimMIM).

• A deep Learning solution for accelerating membrane and nervous fibers detection in human tissues from 3D-Scan image presented by Vincent Blot & Alexandra Lorenzo

Here's a presentation of the Innerve project, the result of a collaboration between AP-HP's CHU de Bicêtre and Quantmetry. This is a medical imaging initiative aimed at revolutionizing the diagnosis of small-fiber neuropathy by making it faster.Since 2017, a team made up of doctors, researchers and engineers has been working on this project, finally culminating in its industrialization phase.Our patented algorithm can process large, complex 3D images, and currently two clinical studies are underway.

If you wish to attend this conference, register now !
🚨 The number of tickets is limited to 50 🚨

We look forward to seeing you 👋