sketch-plot: Progressive Editing for Text-to-Image Academic Figures
2026-06-08 • Human-Computer Interaction
Human-Computer Interaction
AI summaryⓘ
The authors created a tool called sketch-plot to help people easily edit specific parts of academic figures generated by text-to-image models like gpt-image-2. Instead of remaking the whole image to fix one detail, their system breaks the figure into editable pieces that users can adjust as needed. Because computers struggle to separate and vectorize figure parts perfectly, the tool includes a human-in-the-loop step for users to approve or fix how the image is broken down. Their study showed experts preferred using sketch-plot for targeted edits rather than regenerating entire images. This makes editing AI-created academic figures more precise and less frustrating.
text-to-image modelsraster imagevector graphicsimage segmentationhuman-in-the-loopSVGimage editingacademic figuresimage vectorizationinteractive system
Authors
Yinghao Tang, Yupeng Xie, Yingchaojie Feng, Tingfeng Lan, Wei Chen
Abstract
Text to image (T2I) models such as gpt-image-2 can now generate publication grade academic figures from a short prompt, but the output is a flat raster: a user who wants to change one arrow, one label, or one icon has to regenerate the whole image, which also disturbs the parts they wanted to keep. We present sketch-plot, an interactive system that closes this controllability gap with a three layer progressive editing pipeline: a generated PNG, an addressable puzzle of editable pieces, and a per piece SVG. The user stops at the layer that gives them enough control for the change at hand, so the cost of decomposition and vectorisation is paid only on the pieces that need it. Realising this pipeline is not trivial. General segmentation models lack the semantic discriminability to decompose a research figure cleanly, and end to end image vectorisation produces incomplete shapes and loses semantic structure. We therefore route both stages through a human in the loop interface that lets the user accept, refine, or reject decomposition and vectorisation decisions on a piece by piece basis. We validate the design with an expert user study, in which participants found sketch-plot effective for making targeted edits to AI generated academic figures and preferred it over regenerating the whole image. A demonstration video is available at https://anonymous.4open.science/r/SketchPlotVideo/.