# Learning to Segment from Scribbles using Multi-scale Adversarial Attention Gates Source: [Learning to Segment from Scribbles using Multi-scale Adversarial Attention Gates]() Collection index: [llms.txt]() Data updated: 2025-10-25T23:06:39.153Z Authors: Gabriele Valvano, Andrea Leo, Sotirios A. Tsaftaris Published: 2020-07-02T14:39:08.000Z ## Abstract Large, fine-grained image segmentation datasets, annotated at pixel-level, are difficult to obtain, particularly in medical imaging, where annotations also require expert knowledge. Weakly-supervised learning can train models by relying on weaker forms of annotation, such as scribbles. Here, we learn to segment using scribble annotations in an adversarial game. With unpaired segmentation masks, we train a multi-scale GAN to generate realistic segmentation masks at multiple resolutions, while we use scribbles to learn their correct position in the image. Central to the model's success is a novel attention gating mechanism, which we condition with adversarial signals to act as a shape prior, resulting in better object localization at multiple scales. Subject to adversarial conditioning, the segmentor learns attention maps that are semantic, suppress the noisy activations outside the objects, and reduce the vanishing gradient problem in the deeper layers of the segmentor. We evaluated our model on several medical (ACDC, LVSC, CHAOS) and non-medical (PPSS) datasets, and we report performance levels matching those achieved by models trained with fully annotated segmentation masks. We also demonstrate extensions in a variety of settings: semi-supervised learning; combining multiple scribble sources (a crowdsourcing scenario) and multi-task learning (combining scribble and mask supervision). We release expert-made scribble annotations for the ACDC dataset, and the code used for the experiments, at https://vios-s.github.io/multiscale-adversarial-attention-gates ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Vessel Segmentation on OCTA 3M FULL projection 500 (test)](): Vessel segmentation; OCTA 3M FULL projection 500 (test); Dice Coefficient: 76.94. - [Vessel Segmentation on OCTA-500 3M OPL-BM projection](): Vessel segmentation; OCTA-500 3M OPL-BM projection; Dice Coefficient: 72.88. - [Vessel Segmentation on OCTA-500 3M ILM-OPL projection](): Vessel segmentation; OCTA-500 3M ILM-OPL projection; Dice Coefficient: 81.42. - [Vessel Segmentation on OCTA-500 6M OPL-BM projection](): Vessel segmentation; OCTA-500 6M OPL-BM projection; Dice Coefficient: 70.15. - [Vessel Segmentation on OCTA-500 6M FULL projection](): Vessel segmentation; OCTA-500 6M FULL projection; Dice Coefficient: 74.33. - [Vessel Segmentation on OCTA-500 6M (ILM-OPL projection)](): Vessel segmentation; OCTA-500 6M (ILM-OPL projection); Dice Coefficient: 79.48. - [Cardiac Segmentation on ACDC (Dice and HD metrics)](): Cardiac Segmentation; ACDC; Dice (LV): 87.9. - [Medical Image Segmentation on ACDC (test)](): Medical Image Segmentation; ACDC (test); Avg DSC: 81.6. ## Original sources and implementation [Original paper]() [PDF]()