In this talk we will take a deep dive into how we setup and train our guided generation system for our upcoming game Prologue: Go Wayback!. Our guided generation system allows creative people to design rough outlines of levels which then create landscapes which retain natural features. This allows us to generate near-infinite varied landscapes but still include creativity from level designers (and even players!). We will go over data preparation with pysheds, training with hugging face diffusers and shipping to users with ONNX.

Atelier sur la construction d’un modèle de Machine Learning avec Python.

Abstract: Our innate ability to reconstruct the 3D world around us from our eyes alone is a fundamental part of human perception. For computers, however, this task remained a significant challenge — until the advent of Neural Radiance Fields (NeRFs). Upon their introduction, NeRFs marked a paradigm shift in the field of novel view synthesis, demonstrating huge improvements in visual realism and geometric accuracy over prior works. The subsequent proliferation of NeRF variants has only expanded their capabilities, unlocking larger scenes, achieving even higher visual fidelity, and accelerating both training and inference. Nevertheless, NeRF is no longer the tool of choice for 3D reconstruction. Why? Join a researcher from the front lines as we explore NeRF’s foundations, dissect its strengths and weaknesses, see how the field has evolved, and explore the future of novel view synthesis.

Abstract: Invisible image watermarking embeds information into image pixels in a way that remains imperceptible to the human eye but can still be retrieved even after significant image editing. In this talk, after a brief introduction to image watermarking, we’ll explore an approach designed to tackle this issue. Watermark Anything (ICLR 2025) reframes image watermarking as a segmentation problem. We’ll walk through the motivation behind this idea, how we developed and trained the model, the challenges we faced, and the final results. Duration approximately 45 minutes.

This session will showcase AI-powered approaches to detecting, responding to, and preventing cyber threats. Learn how to integrate AI technologies to elevate your cyber defense strategy.

Panel featuring Olena Nahorna, Katarzyna Stoltmann, Jennifer Lapp, Aliya Boranbayeva, moderated by Anita Fechner, discussing AI in communication and data.

AI-powered tools translate complex information into accessible language, supporting data storytelling and collaboration between human insight and machine intelligence. Real-world examples show how organizations use AI to enhance research, reporting, and technical documentation.

The talk explores how large language models (LLMs) have accelerated the development of linguistic features. It focuses on how to adapt feature development processes to match this rapid pace and highlights key considerations for maintaining high-quality output in a fast-evolving AI landscape.

This project focuses on developing a speech-based screening tool for Alzheimer’s Disease (AD) and Mild Cognitive Impairment (MCI) using machine learning models trained on audio recordings from Spanish-speaking individuals. The goal is to create a non-invasive, cost-effective, and easily deployable diagnostic tool for early cognitive impairment detection.

What if everyday objects weren’t so ordinary? In this talk, Alexey will share how he and his son built a GenAI-powered system that transforms photos of common objects into horror stories—automatically generating a new podcast episode every day. He'll dive into the process, deployment, automation, and key lessons learned.

Masterclass to manage end-to-end ML projects using Google Cloud tools. Learn practical strategies to navigate ML implementation, including end-to-end project phases (feasibility, data prep, model training, evaluation, deployment); assess data characteristics and biases; and understand the roles and collaboration needed for ML project teams.

Claims in logistics, especially for lost parcels or delivery issues, can be a significant cost for companies. In this talk, we’ll present the framework and share some early results of a new feature within TrustCourier. This feature uses machine learning to predict and flag high-risk deliveries before they escalate into costly claims.

This talk dives deep into the top AI skills that are shaping the future of the field, drawing insights from cutting-edge research, open-source contributions, industry trends, and job market analysis. From foundational proficiencies in programming, machine learning, and data wrangling to emerging specialties like AI Agents, prompt engineering, and generative AI expertise, we’ll explore what it takes to excel in 2025. Designed for AI practitioners, developers, tech leaders, students, and business professionals, this session offers actionable insights tailored to every audience. This session is your roadmap to the skills and trends that will define the next wave of AI innovation, helping you stay ahead in an ever-evolving landscape.

Overview of AI advancements in data management and robotics; AI-driven automation for data analytics and robotic process automation (RPA); automating data cleaning for visualization and decision-making; strategies for data integration, feature engineering, and AI model deployment; addressing AI bias and ensuring ethical data privacy practices.

A 90-minute hands-on workshop introducing the FiftyOne computer vision toolset. Part 1 covers FiftyOne basics (terms, architecture, installation, and general usage), an overview of useful workflows to explore, understand, and curate data, and how FiftyOne represents and semantically slices unstructured computer vision data. Part 2 is a hands-on session on loading datasets from the FiftyOne Dataset Zoo, navigating the FiftyOne App, programmatically inspecting attributes of a dataset, adding new samples and custom attributes, generating and evaluating model predictions, and saving insightful views into the data.

Explore how data orchestration with Airflow is driving the development of robust, scalable AI solutions.

Explore how data orchestration with Airflow is driving the development of robust, scalable AI solutions and hear directly from those at the forefront of innovation.

DanceBits is an innovative machine learning application designed to automatically segment choreography videos into component moves. Our goal is to create an interactive learning platform for dance instruction and analysis, making dance education more accessible and engaging.

90-minute hands-on workshop on the FiftyOne computer vision toolset. Part 1 covers FiftyOne basics (terms, architecture, installation, and general usage), an overview of useful workflows to explore, understand, and curate data, and how FiftyOne represents and semantically slices unstructured computer vision data. Part 2 is a hands-on introduction to FiftyOne: loading datasets from the FiftyOne Dataset Zoo, navigating the FiftyOne App, programmatically inspecting attributes, adding new samples and custom attributes, generating and evaluating model predictions, and saving insightful views into the data.