Aymen-Lassoued/ABACUS_Bench

ABACUS — A Benchmark for Image Editing Models to Assess Object Category and Counting Under Scene Conservation Instruction-based image editing models can now add, remove, or change a specified number of objects in a scene. Yet they often fail to produce the correct number, and these failures grow sharply as scenes become more complex and the requested count increases. Current benchmarks largely miss this problem: they score edits with holistic similarity metrics or with VLM… See the full description on the dataset page: https://huggingface.co/datasets/Aymen-Lassoued/ABACUS_Bench.