ShieldGemma 2 – Developing safe and responsible AI for images

From the creator

Responsible AI development in multimodal contexts requires robust image safety classifiers. Explore our model for classifying images as "safe" or "unsafe" across key safety categories, which effectively moderates generated images (from any image generation model) and real images (e.g. an image input to a Vision-Language Model), enabling safer use of image generation and VLMs. Subscribe to Google for Developers → https://goo.gle/developers #Gemma #GemmaDeveloperDay Speakers: Dana Kurniawan, Wenjun Zeng Products Mentioned: Gemma

Published 2025-04-02 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.