Doodle Your Keypoints: Sketch-Based Few-Shot Keypoint Detection
This addresses a challenge in machine perception for applications requiring keypoint detection with limited data, though it appears incremental as it builds on existing few-shot and cross-modal methods.
The paper tackles the problem of few-shot keypoint detection when source data is unavailable by using sketches as a source-free alternative, achieving success in few-shot convergence across novel keypoints and classes as demonstrated in experiments.
Keypoint detection, integral to modern machine perception, faces challenges in few-shot learning, particularly when source data from the same distribution as the query is unavailable. This gap is addressed by leveraging sketches, a popular form of human expression, providing a source-free alternative. However, challenges arise in mastering cross-modal embeddings and handling user-specific sketch styles. Our proposed framework overcomes these hurdles with a prototypical setup, combined with a grid-based locator and prototypical domain adaptation. We also demonstrate success in few-shot convergence across novel keypoints and classes through extensive experiments.