Amazon Titan Image Generator v2 AI Model Launched With Improved Image Conditioning Feature

The Amazon Titan Image Generator v2 model can be accessed via Amazon Bedrock.

Written by Akash Dutta, Edited by Siddharth Suvarna | Updated: 7 August 2024 19:00 IST

Amazon Titan Image Generator v2 AI Model Launched With Improved Image Conditioning Feature

Photo Credit: Nvidia

Currently, the Amazon Titan Image Generator v2 is only available in select regions in the US

Highlights

Image conditioning can generate a new image from a reference image
The AI model also allows users to add a preset colour palette for images
Background removal feature has also been upgraded with the new AI model

Amazon announced the launch of its upgraded artificial intelligence (AI) model, Titan Image Generator v2 on Tuesday. Following the release of Image Generator v1 last year, the new image generation model features improved capabilities and functionalities. Aimed at the tech giant's enterprise customers, the AI model can generate images using reference images, edit images, remove backgrounds, and be customised to maintain brand style and subject consistency. Currently, the Amazon Titan Image Generator v2 is available in limited release in select regions within the US.

Amazon Titan Image Generator v2 Features

The company announced the second generation of its enterprise-focused image-generation platform in a blog post. To access it, users in eligible regions will have to go to the Amazon Bedrock console and check model access on the bottom left panel. There, users can request access to the Titan Image Generator G1 v2 AI model.

As per the post, the new model shows a notable upgrade in image conditioning. This feature can generate new images using a reference image and a text prompt. Based on instructions, the generated image can focus on specific visual characteristics such as edges, object outlines, structural elements, and more.

Samsung Wins Patent for a Tri-Fold Smartphone With Specialised Barrier Layer for Improved Durability

The model supports two types of image conditioning. First is Canny edge, which can “extract the prominent edges within the reference image, creating a map that the Amazon Titan Image Generator can then use to guide the generation process.” The second is called segmentation which provides a more granular control over the output as the user can define specific areas within which the AI model can generate new elements.

For brands seeking a consistent approach to their generated images and adherence to the brand colours and design language, the Titan Image Generator v2 offers a colour conditioning feature. Using this, users can specify the colour palette which will be used in the generated images. Users can also add a reference image with the provided hex colours and the AI-generated image will implement the style while adhering to the instructed colours.

Apart from this, the background removal feature has also received upgrades. The company claims it can “intelligently detect and segment multiple foreground objects, ensuring that even complex scenes with overlapping elements are cleanly isolated.”

The Amazon Titan Image Generator v2 is available in the US East (North Virginia) and US West (Oregon) regions. More regions will be added via future updates.

Comments

For the latest tech news and reviews, follow Gadgets 360 on X, Facebook, WhatsApp, Threads and Google News. For the latest videos on gadgets and tech, subscribe to our YouTube channel. If you want to know everything about top influencers, follow our in-house Who'sThat360 on Instagram and YouTube.

Further reading: Amazon, Amazon Titan, AI, AI model, LLM, AI image generator, Artificial Intelligence

Akash Dutta Email Akash Dutta

Akash Dutta is a Senior Sub Editor at Gadgets 360. He is particularly interested in the social impact of technological developments and loves reading about emerging fields such as AI, metaverse, and fediverse. In his free time, he can be seen supporting his favourite football club - Chelsea, watching movies and anime, and sharing passionate opinions on food. More