Google is showcasing its AI advancements at the annual I/O developer conference. The company unveiled new media generation tools, including the enhanced Imagen 4 AI image generator, the Lyria 2 model for producing high-quality audio, and the SynthID detector designed to distinguish AI-generated images from real ones.
Starting with Imagen 4, the image generation tool has been enhanced to create stunning images with impeccable detail like “capturing intricate fabrics, water droplets, and animal fur. It excels in both photorealistic and abstract styles.” Imagen 4 also delivers faster results, supports various aspect ratios, and generates images up to 2K resolution, making them suitable for printing and poster use.
Another major improvement over Imagen 3 is Google’s significant upgrade in spelling and typography. This means no more gibberish text in your creations. The provided image samples clearly showcase how accurate and polished the text generation has become.
If you want to try Imagen 4, it’s available through the Gemini app, Whisk, Vertex AI, and other Google Workspace tools. Google also teased an upcoming, even faster version of Imagen 4 promising to be 10 times quicker than Imagen 3 at generating images. However, there’s no word yet on when this faster model will be released.
Google also introduced Lyria 2, their music generation sandbox that offers a “set of experimental tools to inspire new creative possibilities and help artists explore unique musical ideas.” Lyria 2 is now accessible to YouTube Shorts creators as well as Vertex AI enterprise users.
To protect real-life creators amid this AI boom, Google has launched the SynthID detector a “verification portal that helps people identify AI-generated content.” By uploading a file, you can check if it contains SynthID, a special watermark embedded in creations made with Google’s AI tools.
