Image Answer Analysis — Submitting and Reviewing Visual Answers
PrintThis guide helps students upload, crop, or draw visual answers directly within UnaGPT assignments, and explains how AI auto-grades image-based responses across all educational levels. Each step is accompanied by a screenshot; the red numbered markers in the figures correspond to the numbered instructions in the text.
- Platform: https://app.una.study/
- Applies to: Students (all levels)
Supported image types across all levels
The Image-Answer Analysis feature supports a wide range of visual submission types for primary, secondary, and tertiary levels, including:
- Handwritten diagrams and sketches: science apparatus, design schematics, engineering drafts, and biological illustrations.
- Digital drawings and canvas creations: original artwork or annotations created directly inside UnaGPT.
- Uploaded photos and scans: photos of physical worksheets, hand-drawn graphs, geometry solutions, or mind maps.
Part 1 — Submitting an image answer
When working on an assignment question that requires visual output, locate the Image Answer section below the question text field.
- "OCR (Chinese)" (marker 1): sits with the text answer field above the Image Answer section.
- "Choose Image" (marker 2): uploads an image file from your device.
- "Draw" (marker 3): opens the drawing area to draw an answer directly.
- "Show Instructions" (marker 4): at the top of the page, displays additional learning materials uploaded by the teacher.

- Click the "Choose Image" button.
- Select an image file (e.g. a photo or a scan) from your device.
- The Edit Image modal window will open automatically (fig. 2):
- Cropping: drag the handles of the red selection box to crop out unnecessary background or focus on your answer. (fig. 2 marker 1; fig. 3 marker 1)
- Image controls: use the bottom toolbar to zoom in/out (fig. 3 marker 2), rotate (marker 3), flip horizontally or vertically (marker 4), or move (marker 5) the image.
- Click "Apply" (or "Save") to save the cropped image into your answer box. (fig. 2 marker 2)


- Click the "Draw" button.
- Choose whether to draw on a blank canvas from scratch or annotate on top of an uploaded image.
- Use the drawing toolbar controls (fig. 4):
- Colour palette: pick custom line and highlight colours. (marker 1)
- Pen / eraser: switch tools and adjust stroke thickness. (marker 2)
- Canvas actions: clear the canvas, rotate, zoom, or toggle the question prompt overlay for reference while drawing. (marker 3)
- Save your drawing to attach it to the question. (marker 4)

Part 2 — Modifying or re-submitting answers
- Edit or delete: click "Edit" on an attached image to adjust the cropping or annotations, or click "Delete" to remove it. (marker 1)
- Combining text and image: if required, type additional text explanations or use the OCR button alongside your image submission. (marker 2)
Once satisfied with your entries, click the "Submit" button at the bottom right. (marker 3)

Part 3 — Reviewing AI image feedback
After submitting, UnaGPT evaluates your visual submission against the teacher's grading rubric and provides feedback split into three detailed tabs:
Score Details (marker 1): displays an itemized breakdown of awarded marks based on specific visual criteria. Each row names one criterion with the AI's comment on it (marker 2) and the marks awarded out of the marks available (marker 3).

- Content Analysis: offers line-by-line observations on your drawing, identifying what parts were rendered correctly and what details were omitted.
- Overall Feedback: provides actionable summary comments highlighting key strengths in your diagram and suggestions on how to refine your visual presentation for full marks.
