Hidden-Shot: Towards One-Shot Task Generalization for Low-Level Vision Generalist Models

2026-07-01

AI Summary

Existing general-purpose low-level vision models exhibit limited zero- or few-shot generalization to unseen tasks and lack a systematic evaluation framework. To address this, this work proposes Hidden-Shot, an implicit prompting mechanism that enhances one-shot learning capability in a lightweight manner by extracting implicit task information, incorporating global task-aware textual prompts, and selectively fusing them with intra-task features. We establish the first comprehensive C/U evaluation benchmark for low-level vision—comprising 3C4U and 3C7U protocols—and demonstrate substantial performance gains over state-of-the-art methods across 7 and 10 datasets, respectively. The proposed approach achieves strong generalization to novel tasks without compromising performance on original tasks.

Abstract

Despite the intense engagement surrounding low-level vision generalist models, their effectiveness in zero/few-shot scenarios beyond learned tasks remains unverified. The primary challenge of developing an ideal generalist lies in achieving the ability to generalize from new unseen tasks, which also can be assessed by matched quantitative criteria. Existing methods have made some progress in prompt engineering but have not systematically explored this gap across a wide range of low-level visual tasks. Stimulated by the problem, we propose Hidden-Shot, an implicit prompt mechanism aimed at exploring low-level task adaptation in a vision generalist model. Specifically, the method extracts implicit visual task-based information, utilizes a global task-aware textural prompt, and selectively merges implicit information with in-task processing information to enhance one-shot capabilities in new tasks. The overall design performs direct injection in a cost-effective manner, while minimally altering the architecture of the original generalist model. Additionally, we introduce a data-driven evaluation framework termed C/U assessment to cover two basic scenarios, 3C4U (3 conventional and 4 unconventional tasks) for retraining existing models and 3C7U (3 conventional and 7 unconventional tasks) for training from scratch, as a comprehensive assessment to systematically test the generalization ability of low-level generalist models. Experiments on seven and ten datasets outperform the state-of-the-art vision generalist model, respectively verified by 3C4U and 3C7U framework. Our presented Hidden-Shot approach demonstrates superior performance on one-shot new tasks while maintaining consistent performance on existing tasks.

Problem

• one-shot generalization
• low-level vision
• task generalization
• vision generalist models
• unseen tasks

Innovation

• Hidden-Shot
• one-shot generalization
• low-level vision
• implicit prompting
• task generalist model