Changelog

Updates and improvements to Resemble AI
Back to changelog
March 17, 2026

Visual DFD model update for non-face content

Pushed a visual detection model update focused on content that does not contain faces, in response to a rise in war-related image and video deepfakes on social media. Training added DeepAction, SDFVD, generator sets (Veo3, Kling2, LTX-2), landmark and hard-to-detect image sets (nano-banana-2, nano-banana-pro), and more.