ImageTwin

Imagetwin is an AI-powered software designed specifically for the academic and research community. Its primary function is to analyze images in scientific manuscripts and publications to identify integrity issues. It scans for four main problems: duplicated images within a document, visual manipulations like splicing or copy-move forgeries, plagiarism by comparing against a database of over 120 million published figures, and the presence of AI-generated images, now including model attribution. This helps ensure the visual data supporting research findings is authentic and credible.

The tool works by using advanced image analysis algorithms to examine images in detail. It provides users with confidence scores for each finding, allowing for informed decisions. A key feature is its forensic toolbox, which includes tools like matched keypoints and filters for manual verification. For organizations, Imagetwin offers API access to integrate these checks directly into peer review or publishing workflows, and a private repository feature lets institutions build their own database to auto-check new submissions for internal duplicates. All data is protected with industry-standard encryption.

This tool is most beneficial for academic publishers, research institutions, and individual researchers focused on maintaining high standards of research integrity. Publishers can use it during the manuscript review process to catch errors before publication. Institutions can screen internal research outputs, and individual researchers or integrity consultants, like the cited expert Elisabeth Bik, can use it to efficiently verify the originality of visual data across a high volume of papers, a task that would be incredibly time-consuming to do manually.