Computer Vision Paris Meetup #31

### The Computer Vision Meetup Paris is back 🔥

For this edition, the Meetup will take place on September 24th, at 7:00 p.m at 48 boulevard des Batignolles, 75017 Paris.

🎤 Join us for an exciting evening featuring:

Combining synthetic human data generation and sensor modelling to develop AI-based computer vision solutions presented by Joana Duarte, Computer Vision Engineer | Physics Engineer, PhD at 7 Sensing Software.

Acquiring data for AI is challenging, even more when the subjects of interest are humans, due to the additional ethical, legal, privacy, safety and diversity concerns.
Acquiring real data would be at best hugely expensive, but in newly emerging applications, this is often not even possible, either because the sensors are not available yet, or because the ideal system set-up still needs to be built. For these reasons, synthetic data generation of human subjects arises as a key method to execute multiple iterations on system design and accelerate development for many use cases.
At 7 Sensing Software, we developed a synthetic human face generator combined with sensor modelling to generate large and diverse training datasets. Such pipeline can be customized to fit the targeted use case with the right level of diversity and the required ground truth. 7 Sensing has successfully applied such synthetic human datasets in several use cases such as face recognition, 3D avatar reconstruction, immersive video conferencing and photography enhancement.
In the presentation, we will show how we applied synthetic human data to tackle the problem of eye gaze tracking for Smart AR and VR glasses, and how it helped us to co-design an optimized near-infrared eye-tracking system.

Introducing Lours : empowering pandas for object detection datasets presented by Clément Pinard, PhD, research in deep learning & computer vision

Introducing LOURS. In this talk, we present LOURS (the bear, in French), our new open source library, aimed at helping researchers to easily manipulate datasets. Powered by pandas under the hood, LOURS will let you manipulate an object detection dataset as if it was a simple dataframe.
Cumbersome operations like conversion between dataset formats, splitting or merging are all accessible through a simple yet powerful fully documented API that will let you focus on actual research related tasks.
In addition to dataset manipulation, LOURS offers a powerful evaluation feature which lets you visualize the influence of parameters across your dataset for a specific metric. Finally, this presentation will feature a real world case study where engineers XXII used LOURS in their workflow to train a detection model, evaluate it and elaborate a set of camera and scene specifications for guaranteed quality of detection.

Reserve your spot now ⭐️
We can't wait to see you there.

#ComputerVision #MeetupTech #ParisEvents