Describe what you see
Describing images accurately has been an important part of Cambridge English exams for several decades. This ranges from saying what you see to comparing and even accurately summarising and highlighting visual information. Teachers and coursebooks approach this kind of training often by teaching the relevant vocabulary and then giving exam style exercises. This makes sense but can get tiring. Can AI flip the script?
Maybe it can:
Teachers can help students learn and use the language for describing images and then ask them to re relate images. This is pretty similar to the classic ‘describe the image to your partner and get them to draw it’. The resulting comparison then should show how well the 1st described the image BUT it also tests how well the 2nd understands….and can draw 😉 This version removes the human error part.
LANGUAGES, WOOD & TREES
