Is Visual in-Context Learning for Compositional Medical Tasks within Reach?
This work addresses the challenge of flexible, adaptive vision pipelines for medical tasks, but it is incremental as it builds on existing in-context learning methods with a focus on compositional sequences.
The paper tackles the problem of enabling a single model to handle multiple and new compositional medical tasks without retraining, by exploring visual in-context learning with a focus on training for task sequences rather than individual tasks, and introduces a synthetic compositional task generation engine that bootstraps from segmentation datasets to train models for complex tasks.
In this paper, we explore the potential of visual in-context learning to enable a single model to handle multiple tasks and adapt to new tasks during test time without re-training. Unlike previous approaches, our focus is on training in-context learners to adapt to sequences of tasks, rather than individual tasks. Our goal is to solve complex tasks that involve multiple intermediate steps using a single model, allowing users to define entire vision pipelines flexibly at test time. To achieve this, we first examine the properties and limitations of visual in-context learning architectures, with a particular focus on the role of codebooks. We then introduce a novel method for training in-context learners using a synthetic compositional task generation engine. This engine bootstraps task sequences from arbitrary segmentation datasets, enabling the training of visual in-context learners for compositional tasks. Additionally, we investigate different masking-based training objectives to gather insights into how to train models better for solving complex, compositional tasks. Our exploration not only provides important insights especially for multi-modal medical task sequences but also highlights challenges that need to be addressed.