Skip to yearly menu bar Skip to main content


In-Person Oral presentation / top 5% paper

Universal Few-shot Learning of Dense Prediction Tasks with Visual Token Matching

Donggyun Kim · Jinwoo Kim · Seongwoong Cho · Chong Luo · Seunghoon Hong

AD12

Abstract: Dense prediction tasks are a fundamental class of problems in computer vision. As supervised methods suffer from high pixel-wise labeling cost, a few-shot learning solution that can learn any dense task from a few labeled images is desired. Yet, current few-shot learning methods target a restricted set of tasks such as semantic segmentation, presumably due to challenges in designing a general and unified model that is able to flexibly and efficiently adapt to arbitrary tasks of unseen semantics. We propose Visual Token Matching (VTM), a universal few-shot learner for arbitrary dense prediction tasks. It employs non-parametric matching on patch-level embedded tokens of images and labels that encapsulates all tasks. Also, VTM flexibly adapts to any task with a tiny amount of task-specific parameters that modulate the matching algorithm. We implement VTM as a powerful hierarchical encoder-decoder architecture involving ViT backbones where token matching is performed at multiple feature hierarchies. We experiment VTM on a challenging variant of Taskonomy dataset and observe that it robustly few-shot learns various unseen dense prediction tasks. Surprisingly, it is competitive with fully supervised baselines using only 10 labeled examples of novel tasks ($0.004\%$ of full supervision) and sometimes outperforms using $0.1\%$ of full supervision. Codes are available at https://github.com/GitGyun/visual_token_matching.

Chat is not available.